Spencer v. State
302 Ga. 133
Ga.2017Background
- Spencer was stopped for a nonworking headlight; officer observed slurred speech, odor of alcohol, a wristband, and a cup appearing to contain alcohol.
- Officer administered the horizontal gaze nystagmus (HGN) test; Spencer exhibited four of six HGN "clues."
- At trial the officer testified that, based on his training and experience, four out of six HGN clues "generally indicates a blood alcohol level equal to or greater than .08." Spencer objected and obtained a continuing objection.
- The jury convicted Spencer of DUI (less safe) and possession of an open container; she appealed only the DUI conviction. The Court of Appeals affirmed, distinguishing the officer’s testimony as general rather than a numeric estimate of Spencer’s BAC.
- The Georgia Supreme Court granted certiorari and held the officer’s testimony correlating a specific numeric BAC ("equal to or greater than .08") to HGN results was admitted without the Harper foundational showing of scientific validity/reliability, and reversed the DUI conviction as not harmless error.
Issues
| Issue | Plaintiff's Argument (Spencer) | Defendant's Argument (State) | Held |
|---|---|---|---|
| Admissibility of HGN as proof of impairment | HGN may show impairment but not a numeric BAC; officer lacked scientific foundation to tie 4/6 clues to ≥ .08 | Officer may testify that 4/6 HGN clues generally indicate impairment and generally correlate to BAC ≥ .08 based on training/experience | Court: HGN is admissible to show impairment generally, but testimony correlating specific numeric BAC (or a threshold like ≥ .08) requires Harper foundation and was not shown here; admission was abuse of discretion |
| Whether Harper judicial notice/ foundation was satisfied | Harper requires evidence establishing scientific validity before numeric BAC inference; State presented no studies or expert foundation | State argued prior appellate decisions and officer’s training suffice; testimony was general not a specific estimate for Spencer | Court: State failed to meet Harper; officer’s training and unproduced/unidentified studies were insufficient |
| Distinction between saying "generally indicates ≥ .08" and providing a numeric estimate | Spencer: such language functions as a numeric BAC assertion and is barred without foundation | State: testimony was general, not a direct numeric estimate of Spencer’s BAC | Court: No valid distinction — both present a numeric threshold to jury and require Harper foundation |
| Harmless-error analysis | Admission of the testimony was prejudicial given limited other indicators of impairment | State argued other evidence supported conviction | Court: Error not harmless; reversal of DUI conviction warranted |
Key Cases Cited
- Harper v. State, 249 Ga. 519 (trial judge must determine scientific verifiable certainty before admitting novel scientific evidence)
- Bravo v. State, 304 Ga. App. 243 (HGN not admissible to quantify specific BAC absent scientific foundation)
- Parker v. State, 307 Ga. App. 61 (4/6 HGN clues constitute evidence of impairment)
- Webb v. State, 277 Ga. App. 355 (discussed limits on HGN as basis for numeric BAC)
- Kirkland v. State, 253 Ga. App. 414 (involved breath test results; not a reliable precedent for HGN-based BAC estimates)
- Hawkins v. State, 223 Ga. App. 34 (HGN admissible as evidence of impairment generally)
- State v. Rose, 86 S.W.3d 90 (admission of officer testimony implying a defendant’s BAC crossed legal limit was abuse of discretion)