March-April 2014

Ex parte Coty and the fallout from an unethical lab tech

Emily Johnson-Liu

Assistant Criminal ­District Attorney in Collin County

The jobs we do in prosecuting cases depend heavily on the integrity of the evidence we present in court. When we offer lab results, call an expert to the stand, or even plead a defendant to a charge, we expect those results to be reliable, accurate, and true. And when they are not, the public’s confidence is shaken, and the criminal justice system struggles with how to right those wrongs.
    Just that issue was present in a recent case before the Court of Criminal Appeals, Ex parte Coty.1 In that decision, the court grappled with the fallout from one laboratory technician’s unethical conduct and how far the taint of his behavior should reach.
    In January 2012, a lab technician at the Department of Public Safety (DPS) crime lab in Houston discovered that his coworker, Jonathan Salvador,2 had used the alprazolam (Xanax) from one case to generate the data supporting his identification of alprazolam in another case.3 As a result, the lab report indicating that alprazolam was present in the sample was not based on data from the sample Salvador was supposed to test. The results were based on false or fictionalized data, a practice commonly called “dry-labbing.”
    DPS quickly suspended Salvador and retested all the evidence he had handled in the three months leading up to his suspension. Other errors surfaced, and DPS drew up a comprehensive list of all the cases Salvador had handled over his entire tenure at the lab—six years’ worth of cases. The list included 4,944 cases (9,462 separate pieces of evidence) in 36 different counties in the Gulf Coast region.4 In April 2012, DPS disseminated the list among the various affected DA’s offices, and word spread to the defense bar and some individual defendants. DPS offered to re-test any of the evidence on request, if it had not already been destroyed, and DA’s offices began making choices about how to proceed on the pending cases (which likely involved Salvador’s recent and perhaps more suspect work) as well as those cases that had already been disposed.
    Some of the convicted defendants in the “Salvador cases,” as they came to be called, began filing petitions for habeas corpus relief. The Galveston County CDA’s Office entered agreed findings in several of these cases, recommending that these convictions be overturned.5 In some cases, the drug evidence had already been destroyed and so could not be re-tested. Without the ability to re-test the evidence, no one could be certain that the evidence actually was what Salvador had represented it to be in his report. When these cases reached the Court of Criminal Appeals, the high court verified that Salvador alone had prepared the report in each defendant’s case and that the evidence was no longer available for re-testing. The court then set aside the convictions.6 This was not an outright acquittal, but without the drug evidence, it would be impossible in most of the cases for a prosecution to proceed, so these defendants would likely be released and their cases dismissed.
    Around the same time, the Court of Criminal Appeals overturned the conviction in another Salvador case. But this time, it did so even though the drug evidence still existed and could be re-tested by someone other than Salvador. And the decision was published, which meant it would be precedent and thus govern the law in other cases. The decision, Ex parte Hobbs, reasoned that because the evidence had been in Salvador’s custody and his actions were not reliable, the “custody [of the drug evidence] was compromised, resulting in a due process violation.”7 Other Salvador cases had their convictions summarily set aside in the wake of Hobbs, even where re-testing showed Salvador’s original report had been accurate.8
    Although the court was treating all the convictions in the Salvador cases the same, the State stood a better chance at being able to retry those cases where evidence still existed—these defendants were not necessarily going to be immediately freed. But there was another concern. This very short published case, Ex parte Hobbs, seemed to indicate that any time a lab technician committed misconduct of any kind at any time, that discovery would summarily invalidate all of the cases on which that lab tech had worked throughout his career.
    None of the convictions being overturned involved the 2012 alprazolam cases where Salvador actually committed misconduct by taking evidence from one case and using it to generate data for another. Instead, Salvador’s misconduct was being imputed to the remainder of his body of work, a precedent that could have far-reaching implications on future lab tech errors, potentially invalidating convictions where the role of the lab report was not as central as in a drug possession case.
    A few weeks after issuing another decision in lock-step with Hobbs, the Court of Criminal Appeals took the unusual step of granting rehearing on its own motion. That case was Ex parte Coty, a case from Harris County.9 The court asked the parties to address this unresolved question: Under what circumstances, if any, should the court presume a due-process violation in a case handled by a forensic scientist who has been found to have committed misconduct in another case?10 The defense argued that the presumption that a due process violation occurred should continue in all the Salvador cases because it was too onerous a burden to prove that Salvador had actually committed misconduct in the defendant’s particular case.11 The State argued that Salvador’s misconduct was not severe enough for the court to presume a due process violation and that in every case, the court should consider the importance of the evidence to the rest of the case.12
    When the court issued its opinion on rehearing in Ex parte Coty, however, instead of analyzing the issue as a question of due process as it asked the parties to do, the court analogized the Salvador cases to convictions based on false evidence.13 To prove a false-evidence claim, it is the defendant who must show first, that the evidence in his case was false, and second, that it was material to his conviction or punishment. In a unanimous opinion authored by Judge Hervey, the court held that for cases involving misconduct by a lab technician, the second requirement is the same as for false evidence: The defendant must shoulder the burden of showing that the evidence is material to his conviction. But the court modified the first requirement. Instead of having to prove false evidence by the lab tech in his own case, the defendant can rely on an inference of false evidence if he shows five things:
1)    the lab tech is a state actor;
2)    the tech committed more than one act of intentional misconduct;
3)    the tech worked on the defendant’s case;
4)    the misconduct is the type of misconduct that would have “affected the evidence” in the defendant’s case; and
5)    the tech handled the evidence “within roughly the same period of time as the other misconduct.”
    If the defendant establishes those five elements, he is entitled to the inference that misconduct tainted his own case. But that inference is still rebuttable; the State can rebut the inference if it has evidence that the tech in fact committed no misconduct in the defendant’s particular case.
    After setting out this framework for use in the Salvador cases, the court remanded Coty to the habeas judge to sort out how it would apply. The court also left other issues unsettled. For instance, in repeatedly discussing “evidence” and “false evidence” throughout this test, Judge Hervey never sets out whether “evidence” refers to the physical exhibit (the white, powdery substance) submitted for testing, Salvador’s lab report, or perhaps both. In his concurring opinion, Judge Price states that the “false evidence” is Salvador’s lab report itself.14 For him, the potential problem with the convictions in the Salvador cases is that they may rest upon a lab report in which Salvador falsely claimed to have tested the evidence. And re-testing or other evidence that shows that Salvador’s results were probably reliable still does not speak to whether the lab report was falsified.  
    Others on the court may be concerned with the danger that Salvador may have tampered with the physical exhibits in any given case; thus, it is the physical substance submitted for testing that could be “false.” This was the concern voiced in Hobbs, and in Coty, it appears again, in tandem with the inference that the evidence is false. Judge Hervey’s opinion seems to indicate that by showing the five factors for the inference of falsity, the defendant will thereby also establish that a lab tech’s sole possession of a substance and testing results derived from that possession are unreliable.15 Under this type of false-evidence claim, the State may be able to argue that Salvador’s particular form of dry-labbing would not likely have affected or tainted the sample in the defendant’s case. As unethical as Salvador’s conduct was in creating his lab report in the alprazolam case, if there was never any indication that he tampered with or co-mingled any of the substances submitted for testing, then there is arguably no basis to conclude that any physical evidence has been compromised, and thus it cannot be “false.” There is some indication that this was how the Forensic Science Commission may have viewed Salvador’s misconduct given its statement that “there was no evidence to suggest that there were property control issues of a systemic nature that might preclude future re-testing of evidence.”16
    There is also support for this argument in the court’s opinion. In footnote 11, Judge Hervey wrote that “the scope of the inference of falsity” that Coty showed in his habeas writ “does not appear to extend to a categorical inference of falsity in all respects,” adding that “the inference of falsity should be limited to the pattern of intentional misconduct proven.”17 It may be, for example, that the evidence of Salvador’s misconduct shows that he resorted to pulling a different sample from the drawer only to shore up his documentation when the sample he was faced with analyzing was particularly tricky.18 If so, the State should argue for a narrow inference that goes only as far as that pattern of conduct—e.g., cases where the chemistry was tricky. Cases that did not involve a tricky or difficult analysis might then present a chance for the State to rebut an inference narrowed to Salvador’s particular pattern of misconduct.
    If the inference ends up being that Salvador used data from a different case to support his report in the defendant’s case, the State might still be able to rebut that inference with circumstantial facts, as in Coty, where Salvador did not access any other cocaine samples during the time he was working on Coty’s case. Much will depend on what evidence is presented in the various district courts handling the Salvador cases, and that evidence will likely vary from case to case, at least until the Court of Criminal Appeals considers one of these cases again. So there is potentially still a lot left unresolved.  
    But even if many of the Salvador cases end up with their convictions set aside just as they would have under the Hobbs one-size-fits-all approach, the decision in Coty is still a step in the right direction. The decision has laid the groundwork for how courts can approach the next incident of lab misconduct that may be on the horizon. And this time, the approach will be tailored to the timing and scope of the misconduct and will take into account the importance of the lab work to the overall case. Further, now that we know that the fallout of a lab tech’s misconduct will depend on the depth and breadth of his misdeeds, laboratories will have all the more incentive to discover and disclose misconduct at its earliest stages. And that will serve us all.


1 Ex parte Coty, No. WR-79318-02, 2014 WL 128002, at *1 (Tex. Crim. App. Jan. 15, 2014).
2 The opinion of the Court of Criminal Appeals in Coty states that Salvador worked at the “Houston Police Department’s Crime Lab Division,” but every other resource indicates Salvador worked for the DPS Crime Lab in Houston, and in fact, elsewhere in the opinion the court states Salvador worked for DPS. Ex parte Coty, No. WR-79318-02, 2014 WL 128002, at *1 (Tex. Crim. App. Jan. 15, 2014).
3 Report of the Texas Forensic Science Commission (TFSC), hereinafter “TFSC Report,” dated April 5, 2013, on the Texas Department of Public Safety Houston Regional Crime Laboratory Self-Disclosure, available on the TFSC website,
4 Id. at 9-11.
5 See, e.g., Ex parte Artmore, No. AP-79,982, 2013 WL 831178 (Tex. Crim. App. Mar. 6, 2013) (not designated for publication); Ex parte Eagleton, No. AP-76,986, 2013 WL 831237 (Tex. Crim. App. Mar. 6, 2013) (not designated for publication).
6 See, e.g., Ex parte Turner, 394 S.W.3d 513 (Tex. Crim. App. Feb. 27, 2013).
7 Ex parte Hobbs, 393 S.W.3d 780 (Tex. Crim. App. Mar. 6, 2013).
8 See, e.g., Ex parte Smith, No. AP-76,988, 2013 WL 831359 (Tex. Crim. App. Mar. 6, 2013) (not designated for publication).
9 Ex parte Coty, No. WR-79,318-02, 2013 WL 3250776 (Tex. Crim. App. June 26, 2013).
10 Id.
11 See Applicant’s Brief in Ex parte Coty, 2013 WL 4778514.
12 See Respondent’s Brief in Ex parte Coty, 2013 WL 4013574, authored by Harris County Assistant District Attorney Joshua Reiss.  
13 Ex parte Coty, No. WR-79,318-02, 2014 WL 128002, at *6 (Tex. Crim. App. 2014).
14 Id. at *9 (Price, concurring).
15 Id. at *7.
16 TFSC Report, at 24 (also cited in Ex parte Coty, 2014 WL 128002, at *1).
17 Coty, 2014 WL 128002, at n.11.
18 See TFSC Report, at 6-7, indicating the January 2012 incident of misconduct occurred when Salvador was struggling with analyzing what appeared to be a slow-release alprazolam tablet.