Stacy Ernst v. City of Chicago, 837 F.3d 788

Background

Five experienced female paramedics (Ernst et al.) applied for Chicago Fire Department paramedic jobs, took Chicago’s physical-skills entrance exam, and all failed; overall pass rates: ~98% for men, ~60% for women.
Chicago adopted a physical-skills test developed by Human Performance Systems (Deborah Gebhardt) via a concurrent validity study of volunteer incumbent paramedics and three work-sample exercises.
The case proceeded as two trials: a jury trial on disparate-treatment claims and a bench trial on disparate-impact claims; the jury received an altered instruction on intent that focused on a but-for standard.
The district court found disparate-impact defense satisfied by Gebhardt’s validation study and entered judgment for Chicago; the jury returned defense verdict on disparate treatment after being given the but-for instruction.
On appeal, the Seventh Circuit held the jury instruction misstated the required showing on disparate treatment (intent/motive) and remanded for a new jury trial; it reversed the bench-trial verdict on disparate impact, finding Gebhardt’s study unreliable and not validated under 29 C.F.R. § 1607.14(B)(4).

Issues

Issue	Plaintiff's Argument	Defendant's Argument	Held
Jury instruction on disparate treatment	Plaintiffs argued instruction should require proof Chicago acted with anti‑female intent when creating/using the test	Chicago urged a but‑for individualized standard (would have hired her if she were male)	Instruction misstated law; remanded for new jury trial using magistrate judge’s intent‑focused instruction
Validity of Gebhardt’s concurrent validity study	Study failed regulatory standards: non‑representative volunteer sample, unreliable work‑sample (lift and carry), and unvalidated work samples	Study produced statistically significant correlations for three skills and sufficient reliability for most components	Clear error to credit study; work samples not validated as primary on‑the‑job skills; reversed bench verdict and judgment for plaintiffs on disparate impact
Representativeness / sample selection	Volunteers were above‑average incumbents; combining NYC data did not fix non‑representativeness; self‑selection biased results	Chicago argued volunteers were acceptable and combining NY data prevented inflated cutoff	Sample not shown representative of the relevant labor market; combining NYC data did not cure skew; problem for validation requirement
Evidentiary rulings	Plaintiffs sought admission of evidence (e.g., prior conduct by Gebhardt, training comparisons)	Chicago admitted business records and relied on certain testimony; district court limited some plaintiff proofs	Evidentiary rulings were not an abuse of discretion and were affirmed on appeal (most plaintiff objections waived or undeveloped)

Key Cases Cited

Ricci v. DeStefano, 557 U.S. 557 (discusses disparate treatment and disparate impact distinction)
Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (employer not required to submit to formal validation but courts review job‑relatedness)
Furnco Constr. Corp. v. Waters, 438 U.S. 567 (courts should not restructure business practices)
Matthews v. Waukesha County, 759 F.3d 821 (pattern/practice statistics collateral in individual disparate‑treatment claims)
Gillespie v. Wisconsin, 771 F.2d 1035 (reliability must be established before validity)
Guardians Ass’n v. Civil Serv. Comm’n of the City of New York, 633 F.2d 232 (criticizing circular validation by correlating two unvalidated tests)
E.E.O.C. v. Dial Corp., 469 F.3d 735 (timed physical tests and difficulty relative to actual job can undermine validity)
Lanning v. Southeastern Pa. Transp. Auth., 308 F.3d 286 (cutoff must measure minimum qualifications necessary for successful job performance)