People v. Richmond
2017 IL App (1st) 150642
Ill. App. Ct.2017Background
- On April 9, 2007, three separate attacks occurred in Hyde Park; one victim (C.L.) had anal/vaginal swabs collected that produced two DNA profiles, one matching C.L. and a second with unambiguous readings at 9 of 13 loci.
- Police searched the Illinois DNA database using the nine clear loci and found Richmond matched the alleles at all nine loci; ambiguous loci did not exclude him.
- At trial a DNA expert applied the product rule to allele frequencies and testified the nine-locus profile would occur randomly in roughly 1 in 3.9 trillion Black persons (similarly astronomical odds for other races).
- A jury convicted Richmond of three counts of aggravated criminal sexual assault and one robbery count; sentences were imposed and later adjusted on remand.
- In a pro se postconviction petition Richmond argued trial and appellate counsel were ineffective for failing to obtain or use information about the number of nine-locus matches in the Illinois DNA database to challenge the product-rule statistics.
- The trial court dismissed the petition as frivolous and patently without merit; Richmond appealed and the appellate court affirmed.
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| Whether trial counsel was ineffective for not requesting the number of 9-locus matches in the Illinois DNA database | Counsel was not ineffective; product-rule evidence and expert testimony were admissible and reliable | Richmond: counsel should have sought database counts to undermine the product-rule odds and impeach the DNA evidence | Court: Not ineffective; database counts actually support use of the product rule and counsel reasonably omitted the search |
| Whether appellate counsel was ineffective for not raising trial counsel’s alleged failure on appeal | Appellate counsel exercised reasonable judgment in issues presented | Richmond: appellate counsel should have raised the discovery omission as ineffective assistance | Court: No arguable merit to the underlying claim, so appellate counsel not ineffective |
| Admissibility/reliability of product-rule DNA frequency estimates | Product rule is generally accepted and empirically supported | Richmond: product rule may mislead because loci may be correlated; database matches show problem | Court: Empirical analyses (including database match counts) support product rule; not grossly misleading |
| Whether evidence of thousands of 9-locus "matches" in database would likely lead to exclusion of the DNA evidence | State: counts are consistent with expected number of matches given combinatorics and support product rule | Richmond: showing many database 9-locus matches would undercut the expert’s astronomical odds and likely exclude the evidence | Court: Database match counts (e.g., ~900 in Illinois) comport with expected rates and would not have meaningfully assisted defendant |
Key Cases Cited
- People v. Miller, 173 Ill. 2d 167 (Ill. 1996) (addresses early challenges to the product rule and endorses its general scientific acceptance)
- People v. Hodges, 234 Ill. 2d 1 (Ill. 2009) (standards for surviving first-stage postconviction dismissal on ineffective assistance claims)
- People v. Harbold, 124 Ill. App. 3d 363 (Ill. App. Ct. 1984) (explains the product rule: joint probability equals product of individual probabilities)
- People v. Collins, 438 P.2d 33 (Cal. 1968) (en banc) (warning that the product rule produces erroneous results if events are positively correlated)