UNITED STATES DISTRICT COURT DISTRICT OF MASSACHUSETTS UNITED STATES OF AMERICA v. JAMES HEBSHIE ) ) ) ) ) Cr. No. 02-10185-NG DEFENDANT’S POST-HEARING MEMORANDUM IN SUPPORT OF 2255 PETITION Defendant James G. Hebshie hereby supplements his submissions in support of his 2255 petition to vacate his conviction and his Rule 33 motion for new trial. Hebshie refers the Court to his memorandum in support of his 2255 petition, filed on June 16, 2009 (Document 137), and his reply memorandum, filed on October 16, 2009 (Document 147). I. Introduction On April 15, 2010, the Court issued an order setting the case down for an evidentiary hearing. The Court instructed as follows: Specifically, the Court assumes that evidence and argument will focus on two points: First, whether trial counsel for the defendant was ineffective in failing to request a Daubert hearing, Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), on two types of evidence: (I) the cause-and-origin investigation testimony of Lieutenant Todd Myers, Sergeant David Domingos and Wayne Michael Miller; and (ii) the canine evidence presented by Sergeant Douglas Lynch. Second, whether trial counsel's rebuttal to expert testimony was so deficient as to render his representation ineffective. An evidentiary hearing was held over four days, on June 23, July 9, July 23, and July 27, 2010. The defendant called John Lentini, a pre-eminent expert on arson science and cause-andorigin determinations; trial counsel John T. and John S. Spinale; defendant’s cause-and-origin expert at trial John J. Titus, a fire protection engineer; attorney Peter Muse, who had represented the defendant from November 2004 to March 2005; Daniel Cronin, principal of Phoenix Investigations, the building owner’s insurer’s cause-and-origin investigator; and Professor Jane Campbell Moriarty, University of Akron Law School. The government called two witnesses, Mass. State Police Sgt. David Domingos, who had testified at trial with respect to the cause and origin, and ATF Special Agent Michael Marquardt. The proposed findings and rulings herein focus on the Daubert-Rule 403 issues addressed in the evidentiary hearing. By limiting the proposed findings as indicated, Hebshie does not waive his contentions that counsel was ineffective in failing to object to the evidence of the firefighter’s heart attack and to the erroneous jury instruction on mail fraud, and that there is a reasonable likelihood that the outcome of the proceeding was affected by those errors. II. Proposed Findings of Fact1 and Rulings of Law A. The Applicable Standards: Strickland and Daubert 1. Strickland: The Test for Ineffective Assistance of Counsel In order to show that the defendant was deprived of effective assistance of counsel as guaranteed by the Sixth Amendment, the defendant must establish both that counsel’s performance was deficient and that he was prejudiced as a result. Strickland v. Washington, 466 U.S. 668 (1984); Peralta v. United States, 597 F.3d 74, 79-80 (1st Cir. 2010). First, the defendant must show that counsel's performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the "counsel" guaranteed the defendant by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the 1 To the extent that the facts established at the hearing are not fairly contained within Hebshie’s petition, Hebshie seeks leave of court pursuant to Fed. R. Civ. P. 15(b) to amend the petition to conform to the evidence. Leave should be freely granted. The government is in no way prejudiced by such an amendment, which arises out of the same occurrences as were raised in the original petition. Cf. Mayle v. Felix, 545 U.S. 644 (2005). 2 defense. This requires showing that counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable. Id. at 687. With respect to the first prong, the Court held, “When a convicted defendant complains of the ineffectiveness of counsel's assistance, the defendant must show that counsel's representation fell below an objective standard of reasonableness. . . . Counsel also has a duty to bring to bear such skill and knowledge as will render the trial a reliable adversarial testing process.” Id. at 688. With respect to the second prong, the Court likened it to the standard of materiality of undisclosed exculpatory evidence under United States v. Agurs, 427 U.S. 92, 104, 112-13 (1976): “The defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Strickland, 466 U.S. at 694. A reasonable probability is less than a preponderance. Id. at 693. 2. Daubert and Rule 702: The Test for Admissibility of Expert Opinion Under Rule 702 of the Federal Rules of Evidence, a qualified expert witness may testify “in the form of an opinion, or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.” Rule 702 “imposes a gate-keeping function on the trial judge to ensure that an expert’s testimony both rests on a reliable foundation and is relevant to the task at hand.” United States v. Mooney, 315 F.3d 54, 62 (1st Cir. 2002), citing Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 597 (1993). This gate-keeping function applies not only to “scientific” testimony, but to all expert testimony. Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 147 (1999). The trial court must make the threshold determination as to whether the “expert’s conclusion has been arrived at in a 3 scientifically sound and methodologically reliable fashion.” Ruiz-Troche v. Pepsi Cola of Puerto Rico Bottling Co., 161 F.3d 77, 85 (1st Cir. 1998). The proponent of the evidence bears the burden of proof by a preponderance of the evidence. Daubert, 509 U.S. at 593 n.10. In determining the admissibility of expert testimony, the trial court may consider (1) whether the theory or technique can be and has been tested, (2) whether the expert’s technique or theory has been subject to peer review and publication, (3) the known or potential error rate, (4) the existence and maintenance of standards and controls, and (5) whether the technique has gained general acceptance in the relevant scientific community. Id. at 593-94. These factors are not definitive or exhaustive, and the trial judge may consider other factors to evaluate reliability. Mooney, 315 F.3d at 62. “The issue is whether (a) the opinions and conclusions of the expert are accompanied by information that enables the factfinder to evaluate the likely accuracy of the expert’s opinion, and (b) the information is presented in such a way that the factfinders will not be fooled into excessively overvaluing the testimony.” United States v. Green, 405 F. Supp. 2d 104, 119 (D. Mass. 2005), citing Michael Saks, The Legal and Scientific Evaluation of Forensic Science (Especially Fingerprint Expert Testimony), 33 Seton Hall L. Rev. 1167 (2003) [internal quotations omitted]. “The issue is not whether the field in general uses a reliable methodology, but the reliability of the expert’s methodology in the case at bar, i.e., whether it is valid for the purposes for which it is being offered, or what the Court has described as a question of ‘fit.’” Green, 405 F. Supp. 2d at 119, citing Daubert, 509 U.S. at 591. B. Background and General Observations about Trial Counsel’s Representation of the Defendant in this Case 1. The defendant was represented at trial by John T. (“John”) Spinale, who filed an 4 appearance in March 2005, approximately fifteen months before trial. Tr. Day 2, 126, 128.2 He was assisted by his son John S. (“Jay”) Spinale. Tr. Day 2, 170. They succeeded Peter Muse, who had taken over the case from his father Robert Muse in November 2004. Tr. Day 3, 214-15. Jay Spinale, who handled the forensic evidence aspect of the defense (Tr. Day 2, 174), had never before represented a criminal client in federal court. Tr. Day 2, 171. His solo practice centered primarily on state criminal practice, often in juvenile court, as well as some family law. Tr. Day 2, 170. Like his son, John Spinale, a long-time general practitioner, had limited federal court experience. Tr. Day 2, 123-25. Neither had ever tried a case to a jury in federal district court. Tr. Day 2, 125, 171. 2. Their lack of federal criminal experience was reflected in their conduct of the defense. They filed no pretrial motions except to obtain incidental relief from conditions of pretrial release. Tr. Day 2, 128. They filed no motions in limine, no motion to sequester witnesses, and did not seek a Daubert hearing to challenge the government’s forensic evidence. Id. They filed very few requests for jury instructions, and in support cited Massachusetts rather than federal cases. Proposed Jury Instructions, Document 76. As noted in the Court of Appeals’s decision in Hebshie’s direct appeal, defense counsel failed to object to the Court’s clearly erroneous jury instruction on the elements of mail fraud. United States v. Hebshie, 549 F.3d 30, 42-43 (1st Cir. 2008). Counsel elicited a sanctions warning from the Court for failure to use the mandatory electronic filing system. Docket Entry 85. 3. While there is no requirement that defense counsel have prior experience in 2 References to the transcript of the evidentiary hearing held on the 2255 motion are designated “Tr. Day [#], [page #]. References to the trial transcript are designated “Tr. Tr. [vol. #], [page #]. Trial exhibits are referred to as “Tr. Ex. [#]; hearing exhibits are referred to as Ex. [#]. 5 federal court, counsel in this case coupled inexperience with a failure to become familiar with the rules and case law applicable to federal criminal proceedings. See, e.g., Motion for New Trial, Document 91, and Docket entry, August 7, 2007 (claim of newly discovered evidence is “extraordinary” and “[a]t best, it bears on the effectiveness of counsel . . .”). 4. Jay Spinale testified that he spent “thousands” of hours preparing for trial in this case. Tr. Day 2, 177. This claim strains credulity to the breaking point given the length of time the Spinales were responsible for Mr. Hebshie’s defense.3 In contrast, he spent little or no time preparing to testify in this proceeding. Tr. Day 2, 203-04. While he testified that he had “memorized and knew” the 2004 edition of NFPA 9214, the National Fire Protection Association’s manual for fire investigators, during trial preparation, he could not recall the substance of any of NFPA 921 by the time of the hearing in this proceeding. Tr. Day 2, 177. His testimony was marked by repeated assertions that he could not recall salient details of the case and his preparation. See, e.g., Tr. Day 2, 180, passim. C. Counsel Was Ineffective in Defending the Cause-and-Origin Evidence 5. Attorney Peter Muse, a member of the Massachusetts bar since 1982 and an experienced criminal defense lawyer and long-time member of the Criminal Justice Act panel in 3 To bill a total of 2000 hours in a single year is a difficult task, requiring 40 billable hours each and every week for 50 weeks a year. For a sole practitioner juggling numerous cases as well as the business side of practicing law, it is practically impossible to devote “thousands” of hours to a single case over a 15-month period. 4 While the government repeatedly referred to the 2001 version of NFPA 921 (Ex. 6), that edition is irrelevant to the question of whether the government’s 2001 cause-and-origin investigation passed muster in 2006. The question is not, as the prosecutor would have it, what the prevailing best practices were at the time of the investigation (see Tr. Day 1, 74), but whether in 2006, defense counsel had a basis to challenge the validity and reliability of the opinion evidence. In any event, the two editions are not materially different for the purpose of this case. 6 this district (Tr. Day 3, 213-14), represented the defendant for the four-month period ending in March 2005 when John Spinale filed his appearance. Tr. Day 3, 217. Muse had defended arson charges in two previous cases, and had challenged the forensic evidence in those cases. Tr. Day 3, 215. 6. At the time the Spinales took over the case, fire protection engineer John Titus had already been retained, and Michael Higgins, a chemist, had either been retained or was about to be. Tr. Day 3, 215. In a letter to Robert Muse, Higgins noted that light petroleum distillate (LPD), which the government claimed was used to accelerate the fire, is commonly found in many products and the laboratory had been unable to identify precisely what the substance was. Ex. 17. 7. Muse went over the entire file with Jay Spinale, piece by piece, during a Saturday morning meeting in Muse’s office in Quincy. Tr. Day 3, 218. Muse turned over the Higgins letter to Spinale (Tr. Day 3, 218), who did not recall ever seeing it. Tr. Day 2, 184. The Spinales kept Titus on to assist with the cause-and-origin evidence, but did not retain a chemist to assist with trial preparation or to testify at trial. Tr. Day 2, 132. 8. In 2005-06, when the Spinales were preparing to defend this case, it was the prevailing practice among the criminal defense bar to attempt to exclude or limit the admissibility of technical or scientific forensic evidence in cases where a determination of criminal responsibility rested heavily on the reliability of such evidence.5 Daubert challenges to 5 Jane Moriarty, a professor at the University of Akron School of Law, testified to the standard of performance expected of a criminal defense lawyer with respect to challenging forensic evidence in the relevant timeframe. Tr. Day 3, 327, et seq.; Ex. 14. Professor Moriarty served as a criminal defense attorney at the Federal Defender’s Office in the Western District of Pennsylvania and also practiced law in Boston. Her knowledge of criminal defense practice is also informed by her experience as a judicial clerk, both in Massachusetts, where she clerked for 7 the reliability or accuracy of the technology, in general or in the particular case, motions in limine to exclude the evidence or to limit its admissibility, and voir dire hearings to test the witnesses’ expertise were all means then available to defense counsel and commonly utilized. Attorney Muse anticipated that he would have filed a Daubert motion or a motion in limine in some form to challenge the testimony of the government’s experts. Tr. Day 3, 216. 9. Daubert challenges were being raised regularly in criminal cases in 2005-06. Tr. Day 3, 339. In particular, the limits of expertise based on experience had been a concern at least since 1999 when Kumho Tire was decided. Indeed, the subject was addressed in the Advisory Committee notes to the 2000 amendment to Fed. R. Evid. 702. Tr. Day 3, 340. Moreover, in the period up to the time of trial, there were treatises available to defense counsel that addressed Daubert concerns in the context of arson. Tr. Day 3, 336. See, e.g., Fagan, et al., Modern Scientific Evidence 636 (2002). The pertinent edition of Fagan raised the issue of challenging so-called accelerant-sniffing canines as well as the problem of experience- and anecdotally-based opinions about the cause and origin of fires. Id. at 636; Tr. Day 3, 337-38. While Daubert had not been rigorously applied in criminal cases,6 it was nonetheless incumbent on defense counsel to seek to exclude the government’s opinion evidence if there was a factual basis to do so. Hon. William Young when he was an associate justice of the Superior Court, and in Pennsylvania, where she clerked for the late Chief Justice Cappy of the Supreme Court of Pennsylvania. Tr. Day 3, 328. Her academic areas of interest are evidence, and in particular, expert evidence, and legal ethics, and in particular, the intersection of legal ethics and scientific and expert evidence. Tr. Day 3, 329. She is the author of a two-volume treatise entitled Psychological and Scientific Evidence in Criminal Trials and the co-author of a casebook entitled Scientific and Expert Evidence. She has written a number of articles on forensic science and expert testimony. Id. 6 See generally National Research Council, Strengthening Forensic Science in the United States: A Path Forward 9-13 (2009). 8 10. There were ample factual grounds to challenge the government’s opinion evidence in this case, and the Spinales were aware, or should have been aware, of those grounds. In addition to the Higgins letter, John Titus expressed to Jay Spinale his strong disagreement with the government’s cause-and-origin determination. 11. The defense theory of cause and origin was that (1) the photographs and other evidence were consistent with the fire having started in the basement, most likely originating in electrical wiring; (2) the government’s theory of cause and origin did not explain the phenomena observed by firefighter Todd Myers at the scene or the physical evidence left behind by the fire, in particular, the hot spots on all four walls of the shop and the fire’s resurgence after it was extinguished at the supposed site of origin; and (3) the presence of LPD at the supposed site of the origin of the fire proved nothing in the absence of proof that the LPD was not incidental to the premises. 12. Notwithstanding Higgins’s and Titus’s warnings that the government’s case was not sound, the Spinales did not move to exclude the opinion evidence of cause and origin under Rule 702. Tr. Day 2, 128. Jay Spinale testified that he did not seek a Daubert hearing on the cause-and-origin testimony because he concluded that it was sufficiently reliable. Id. at 189. John Spinale’s testimony reflected confusion about the applicability of Daubert to arson cases. Id. at 160. The Spinales’ rationale for not seeking a Daubert hearing was not reasonable and therefore was not competent. 13. A Daubert challenge should have been undertaken in this case because: (1) the scope of the witnesses’ experientially-based expertise was questionable; (2) Domingos’s investigation fell woefully below the industry standard; 9 (3) there was ample reason to conclude that the cause-and-origin opinions were not scientifically valid, that is, that Domingos and Miller’s opinion did not reflect a sound application of scientific principles to the situation at hand; (4) the complete absence of documentation of the condition of the basement, in light of the defense expert’s opinion that the fire had to have started in the basement, strongly suggested that Domingos and Miller’s opinion was not grounded in sufficient facts and data; (5) the laboratory results indicating the presence of LPD in a single sample of carpet were insignificant without further identification of the substance and proof that the substance was neither incidental to the premises nor the result of pyrolysis; and (6) the relevance of the canine evidence in light of the laboratory sample was minimal, and the risk of prejudice was substantial. See Tr. Day 1, 64-65; Tr. Day 3, 353. 14. Having failed to challenge the government’s opinions pre-trial, defense counsel ineffectively challenged the government’s case at trial. Counsel did not attempt to limit the scope of the government’s opinion evidence to valid, reliable, and relevant areas of inquiry that lay within the scope of the witnesses’ respective areas of expertise. Counsel also failed to object to testimony that clearly was irrelevant, beyond the scope of the witnesses’ expertise, or was unduly prejudicial. Finally, the defense was essentially abandoned in the defendant’s closing argument in which counsel conceded that the jury did not need experts to assist them in understanding how fire works and implored them to use their “common sense.” Tr. Tr. VI, 198. 15. Todd Myers, a firefighter, testified without objection to the significance of white spots found on all four walls by the infrared camera. Tr. Tr. I, 106 (“Some white was showing up on the thermal imaging camera, but we just associated that with areas in the room that were 10 heated by the flames that were just there prior.”); Tr. Tr. II, 5. Not only was this testimony incompetent, it was wrong. Tr. Day 1, 46-47; Day 3, 285, 292-94. Had the walls been warmed by fire located where the government says the fire was located, on the left wall of the newsstand, there would not have been thermal imaging hot spots on all four walls. Tr. Day 3, 285. If the only source of the heat picked up on the thermal imaging camera was the fire on the B wall of the newsstand that Myers had extinguished before the camera was deployed, the spots would not have been white. White spots, properly interpreted, indicate a degree of heat that is much too hot to permit a person to withstand. Tr. Day 3, 292-93. As John Lentini testified, “He said all of the walls were hot, and that suggests that the fire – there was heat behind the walls, again, coming up from the basement rather than heat attaching to those walls from the room because he would have been in the middle of that heat.” Tr. Day 1, 46. The only interpretation consistent with the evidence is that there was fire inside all four walls, and the only logical deduction one can draw is that the fire Myers saw in the rear of the newsstand was not the origin of the fire, but rather breakout fire from the walls which had come up from the cellar. That the firefighters “might have” poked holes into the walls (Tr. Day 3, 311), even assuming they did what they “might have” done, indicates only that the fire had already reached the upper floor. As Titus explained, “[F]ire travels extremely fast. It goes up the wall like a chimney. It’s going to move a lot faster than you or I can, of course, and it’s got a head start.” Tr. Day 3, 293. 16. If Myers’s testimony had been challenged during trial on the basis of his lack of credentials to offer such an opinion or the scientific validity of the testimony, it would have been excluded. 17. The means were available to defense counsel at trial to create a substantial doubt 11 that Domingos had ever gone into the basement to investigate whether the fire started there. Defense counsel did not adequately exploit the complete absence of photographs of the basement.7 Nor did defense counsel make use of the report of investigation prepared by Phoenix Investigations, which had been hired by the building owner’s insurer to investigate the cause and origin of the fire.8 Phoenix’s business records document that on Monday morning, April 23, the day after Domingos’s site investigation, Domingos requested Phoenix’s principal Dan Cronin to contact him if he was going to the site because Domingos wanted to get into the basement. Tr. Day 3, 233; Ex. 8. Cronin, who called but did not reach Domingos to coordinate a return visit, went to the site with only his own personnel. Tr. Day 3, 234-36. Like the government agents, Cronin took no photographs of the basement. Tr. Day 3, 237; cf. Ex. 9 [the Phoenix photographs]. He testified, credibly, that, although it is usual practice to photograph the basement of a building, he did not do so because it was impossible to get in the basement. Id. His site drawing, while far, far more detailed than the government’s drawing (Tr. Ex. 7), does not depict the location of the stairwell leading to the cellar from the newsstand because Cronin could not see the stairwell. Tr. Day 3, 239; Ex. 10. Domingos admitted that he never went back to the site after April 22 for further investigation. Tr. Day 3, 409. D. Defense Counsel Was Ineffective in Failing to Move to Exclude or Limit the Canine Testimony 18. At trial, Massachusetts State Police canine handler Doug Lynch testified at great 7 Two photographs were taken of the stairwell into the cellar, both from the vantage point of the first floor. Tr. Ex. 12G; Ex. 13. 8 Jay Spinale did not seek to obtain further documentation of the condition of the basement and did not contact Daniel Cronin, principal of Phoenix Investigations, or any of the insurance investigators to obtain additional photographs. Tr. Day 2, 191. 12 length about his dog Billy’s alert along the B wall of the newsstand. In closing, the government argued that the dog’s failure to alert elsewhere indicated that the alert occurred at the site of the origin of the fire. Tr. Tr. VI, 171. The Lynch testimony, buttressed by the closing argument, was powerful and effective. 19. Jay Spinale testified that he did not seek to exclude the canine evidence because “every case in every jurisdiction, every case that I read prior to this trial, every jurisdiction permits the canine and to the point where they permit canine alert testimony even when there’s a negative sample.” Tr. Day 2, 178; see also id. at 136. Spinale admitted that there was no strategic reason not to seek a Daubert hearing with respect to the canine, and that he would have objected to or moved to exclude the government’s expert testimony if he thought there was a basis for excluding it. Id. at 168. 20. In fact, there was a basis to exclude, or at a minimum, limit the canine evidence. NFPA 921 makes clear that canine handler teams are to be used only to assist in the location and collection of samples for laboratory analysis for the presence of ignitable liquids. NFPA 921, ¶ 16.5.4.7, Ex. 2. NFPA’s position that canine teams are useful only as an investigative tool is consistent with research findings that dogs are prone to false positives. See, e.g., Kurz, Michael E., et al., Evaluation of Canines for Accelerant Detection at Fire Scenes, 39 Journal of Forensic Sciences 1528 (1994) (Ex. 12); Katz, Sheryl R., et al., Unconfirmed Canine Accelerant Detection: A Reliability Issue in Court, 43 Journal of Forensic Sciences 329 (1998). Even ATF Special Agent Michael Marquardt, the government’s expert witness at the hearing, conceded that canines are nothing more than a tool for site investigation. Tr. Day 4, 515. 21. Substantial questions about the reliability of canine evidence had indeed been 13 raised in the courts, including the U.S. Supreme Court, by 2005. See Illinois v. Caballes, 543 U.S. 405, 411-12 (2005)(Souter, J., dissenting)(“The infallible dog, however, is a creature of legal fiction. . . . [T]heir supposed infallibility is belied by judicial opinions describing well-trained animals sniffing and alerting with less than perfect accuracy, whether owing to errors by their handlers, the limitations of the dogs themselves, or even the pervasive contamination of currency by cocaine. . . . Indeed, a study cited by Illinois in this case for the proposition that dog sniffs are "generally reliable" shows that dogs in artificial testing situations return false positives anywhere from 12.55 to 60% of the time, depending on the length of the search.”), and cases cited therein.). 22. If counsel was not already on alert to the limitations of canine evidence by virtue of NFPA 921 and articles in the forensic literature raising concerns about the lack of data and standards to support claims of canine accuracy and reliability (Tr. Day 3, 342-43), this Court’s repeated inquiries as to whether the dog evidence would be subject to a Daubert hearing should have prompted counsel to look at the question more closely. It did not. See Tr. Tr. III, 60, 121; IV, 112. 23. Counsel could not recall asking the government for reports on the testing of the dog at issue. Tr. Day 2, 141. There is no indication that any such reports were produced to the defense. At a Daubert hearing, the burden would have been on the government to establish the reliability of the dog to a fair preponderance of the evidence. 509 U.S. at 593 n.10. While the accuracy of the alert with respect to the single sample taken is arguably demonstrated by the positive laboratory report, that begs the question of what, if anything, the unsubstantiated and subjective testimony about the dog’s accuracy added to the government’s case. The answer is 14 twofold: first, that the dog was alerting, not merely to a hydrocarbon, but to an “accelerant”; and second, that the dog’s failure to alert elsewhere was used to pinpoint the location of the alert as the point of origin. As to these uses, there was no proof of the dog’s accuracy, and there could be no proof. Had defense counsel challenged the use of the canine non-alert in a Daubert hearing or sought to prevent the prosecutor and witnesses from claiming that the dog could distinguish between pyrolytic products and “accelerants”, the government would not have prevailed. 24. In what is perhaps the most egregious of many serious lapses at trial, counsel did not object under either Rule 702 or Rule 403 when Lynch began testifying far beyond the scope of the arguably appropriate use of a detection canine to narrow the search for samples to be taken for laboratory testing. Tr. Day 2, 183. The jury was treated to extensive irrelevant and inflammatory testimony about Billy, the canine at issue, and to an unrelated dog, Amore, who had also been trained by Lynch. The testimony about the dogs extended over 56 pages of direct examination. Tr. Tr. III, 139-179; IV, 8-24. Much of the ostensibly expert testimony was entirely subjective and unreliable, e.g., testimony that the dog gave a “strong” alert, a “solid hit.” Tr. Tr. IV, 17. Counsel also did not object to repeated references to the dog as an “accelerant detection canine” (e.g., Tr. Tr. III, 141, 145, 157) even though the dog is trained to alert to hydrocarbons, not “accelerants.” Counsel also failed to object to or in any way challenge Lynch’s unsubstantiated, ipse dixit testimony that Billy was 97% accurate (Tr. Tr. III, 156)9, and to testimony as to which no expertise was demonstrated, e.g., how the “olfactory system” of the 9 There are no peer-reviewed standardized methods of training detection dogs and their reliability is highly variable. Tr. Day 3, 306. By the time this case went to trial, the Journal of Forensic Science had reported the results of studies indicating a substantial error rate with canine alerts, far higher than Lynch asserted with respect to his dog. Tr. Day 3, 343; see, e.g., Ex. 12. 15 dog works, why it is important to the dog (Tr. Tr. III, 141), what is instinctual in a dog (Tr. Tr. III, 142), etc., etc. Lynch’s testimony, which was the centerpiece of the government’s case, was often very far afield, including a lengthy explanation of the training of Amore, a narcotics dog who died in 1995, six years before the fire at issue. Id. at 144. The subjectivity of the dog alert was made a centerpiece of the government’s presentation, as if the relationship between dog and handler enhanced the reliability and probative value of the lab results, rather than undermined it: Her uniqueness, there was a lot of uniqueness to her alert, it was beyond the odor, sit, food. There were certain things that over the period of time that I handled this dog, I read her face. I was with her 365 days a year. It was the first thing I did every morning, it was the last thing I did every night for the entire time that I had that dog. I knew her personality, I knew the way her eyes shifted. I knew the ways her ears shifted when she located stuff. There were things she did. Her head would cock to the side, her ears, I don’t know if anyone owns a Labrador. Their ears are kind of expressive. Her ears would go up in the corners, her head would tilt to the side, and her tail would start to got [sic]. There were very distinctive things that she did at the fire scenes that I didn’t see at other places with her. It was an excitement in the dog that she had located what she was looking for. Tr. Tr. III, 150-151. None of this obviously objectionable testimony incited an objection from the defense. 25. Lynch was also allowed to explain, without any demonstrated basis and without objection, the cause of Billy’s few errors. Id. at 156. While admitting that the dog was not infallible, Lynch opined that he himself may have been at fault for failing to collect the sample properly. Id. In short, Lynch’s testimony about Billy’s uniqueness, her specialness, and her prowess at detecting “accelerants” went far beyond any plausible, if marginal, relevance about use of the dog to help select a location for collection of a sample. Defense counsel, while 16 concluding that the testimony, which lasted “like half a day” (it actually extended over parts of two days), was “over the top”, nonetheless did not object because he believed that the testimony was admissible. Tr. Day 2, 148. This assessment was unreasonable and incompetent. 26. The government elicited testimony from Lynch that the dog did not alert to anything else on the premises. Tr. Tr. III, 177; IV, 17, 20. Defense counsel did not object. A dog’s failure to alert was never shown to be scientifically significant in general or in this case. The testimony was not admissible. Counsel should have objected and the failure to do so was unreasonable and incompetent. No viable strategy could justify the failure, and, in any event, the Spinales did not claim the failure was strategic. Had the objection been made, it would have been sustained. 27. The entire canine evidence was grossly out of proportion to any legitimate evidentiary value it arguably had. Even if the dog’s alert added anything of scientific value to the lab results – which it emphatically did not – the dog-related testimony and argument in this case should have been excluded under Fed. R. Evid. 403. Repeated references to the dog’s personality, her demise, and Lynch’s obvious love for the dog injected an emotional quality to the evidence that gave it far more significance than it warranted. In short, the canine evidence was grossly and unfairly and improperly prejudicial. Had the defense mounted any sort of challenge to it, it would have been excluded or severely curtailed. 28. In closing argument, the government argued that the dog’s failure to alert elsewhere on the premises proved that the origin of the fire was located where the dog did alert. “Billy knew what she was smelling for. She indicated an alert on one place and one place alone. It was a strong alert. If it was the carpet that was setting her off, then she would have triggered 17 everywhere else around her.” Tr. Tr. VI, 171. Defense counsel did not object and failed in the defense closing to respond to the government’s argument. Id. at 189. Like the failure to object to the improper canine testimony, counsel’s failure to object to this improper argument was unreasonable and incompetent. Had he done so, the objection would have been sustained. E. Counsel Was Ineffective in Failing to Move to Exclude the Laboratory Analysis of the Carpet Sample where No Comparison Sample Had Been Taken 29. Jay Spinale, notwithstanding his claim of intimate familiarity with NFPA 921, did not challenge the admissibility of the test results performed on a single laboratory sample on the ground that the results in this case, that is, results indicating the presence of an unidentified LPD in a newsstand, are meaningless in the absence of a comparison sample. Without such a comparison, it is not possible to determine whether LPD found in a charred carpet sample was incidental to the pyrolysis of the carpet, was part of the make-up of the carpet itself, or was a foreign substance. Tr. Day 2, 194. 30. Titus testified that he alerted the Spinales to the need for a comparison or control sample in order to ensure that the presence of a chemical in one location is not also present elsewhere. Tr. Day 3, 268.10 31. Lynch testified at trial that he did not take a comparison sample in this case and that he does so in less than 10% of cases. Tr. Tr. III, 169. He also testified that he decided whether to take a comparison sample based on the “type of alert” that he had from the dog, how 10 As noted at the hearing, presence of the same hydrocarbon elsewhere may or may not be consistent with its use as an accelerant, depending on the nature of the hydrocarbon found and other factors that may explain why the substance would or would not be expected to be in multiple locations. To the extent that the test results are relied upon to show point of origin, of course, presence of the substance in multiple locations tends to disprove that claim. 18 many alerts he had from the dog, and whether he was requested to take a sample. Id. at 169-170. In contrast, NFPA 921 indicates that a comparison sample should be taken wherever possible since the need for a comparison sample is made by the lab and not on site. Ex. 2, ¶¶ 16.5.4.6.1., 15.5.4.6.3. Lynch’s claim that he is competent to determine whether a comparison sample is necessary is preposterous. There is no basis in established science to support Lynch’s opinion that the type or number of the dog’s alerts is relevant to the decision whether to take a comparison sample. The need for a comparison sample does not relate to the accuracy of the dog, but to the significance of the laboratory analysis to the investigation. There was no evidence that Lynch was qualified to testify about the chemistry of pyrolysis (Tr. Tr. III, 169), or that “I didn’t have a lot of problems with pyrolysis products with the dog.” (id. at 170), or to describe styrene and its prevalence in the environment. Id. at 171. The scope of Lynch’s expertise with respect to fire and hydrocarbon chemistry could and should have been vetted in a Daubert hearing, or at a minimum, challenged at trial. Counsel’s failure to do so was unreasonable and incompetent. Had a challenge been made, it would have been successful. 32. The government offered no other explanation for failing to take a comparison sample. No showing was made that it was impossible or impracticable to do so. Photographs introduced at trial, Tr. Ex. 12C and 12E, show that the “carpet is in very good shape for a carpet that supposedly had ignitable liquid put on it.” Tr. Day 1, 40. 33. Without a comparison sample, the laboratory results that LPD was found in the carpet was not probative. The results do not permit an inference that the fire started at the location of the alert and they do not permit an inference that the fire was incendiary. The mere presence of LPD is proof of precisely nothing other than the presence of LPD, an irrelevant fact 19 unless the presence is shown to be foreign to the environment. There was no such proof at trial. Because of the risk that the jury would attach weight to this meaningless result, competent defense counsel should have challenged the admissibility of the laboratory sample before trial. Failing that, admissibility should have been challenged at trial. Defense counsel did neither. Instead, counsel weakly criticized the failure to obtain a comparison sample in crossexamination. Tr. Tr. IV, 51-52. This was no substitute for excluding the evidence entirely. F. Counsel Was Ineffective in Conceding Miller’s Qualifications and in Failing to Rehabilitate the Expertise of Titus 34. The Spinales did not utilize Titus to assist them in preparing to cross-examine the government’s experts after they had testified on direct examination. Tr. Day 3, 272. In particular, Titus was not asked to stay to listen to the testimony of the government’s rebuttal witness Wayne Miller. Id. 35. Defense counsel conceded Miller’s qualifications as a cause-and-origin expert, without having asked for a voir dire hearing on his qualifications, the opinions he would testify to, and whether those opinions fell within the scope of any field of expertise. Tr. Tr. VI, 105. While this may have been appropriate in some cases, it was not appropriate strategy in a case that rested so heavily on the whether the experts’ reasoning and application of scientific principles were sound and reliable.11 36. Miller was a private consultant who had been a special agent with ATF for most of his career. Tr. Tr. VI, 103-04. Like Domingos, he was a certified fire investigator. Id. at 108. 11 Mr. Spinale testified at the hearing that he believed that Miller testified the day following Titus’s testimony, which would have afforded him an opportunity to discuss Miller’s qualifications with the prosecution. In fact, Miller testified on the same day as Titus. Compare Tr. Day 2, 154 with Tr. Tr., VI, 103. 20 He had never been to the site. He formed his opinion based on the documentation and photographs of the scene and on an interview with Domingos. Id. at 109, 113. Miller opined, contrary to Titus, that there was no evidence the fire had started in the basement. Id. at 112, 127. Of course, such an opinion is worthless unless it is based on sufficient facts and data. Otherwise, the jury is likely to equate lack of evidence with evidence to the contrary. 37. Miller claimed that the hearing grate showed no evidence that it had been affected by heat. Tr. Tr. VI, 127- 29. There is no indication that Miller was competent to testify as to how heat and fire affect steel. The failure to challenge Miller’s expertise was prejudicial in that he was permitted to testify, without objection or rejoinder, that the condition of the metal did not exhibit any evidence of contact with fire or heat. This testimony went to the jury without any challenge whatsoever. In fact, as Titus would have testified had he been called, the grate as depicted in Tr. Exs. 12C, D, and E, showed clear evidence of damage by heat, including discoloration of the metal and severe rusting, which occurs when steel is exposed to heat and then moisture. See Tr. Day 3, 274-77. The flooring around the grate also exhibited heat damage. Tr. Day 3, 275. According to Titus, had he been given an opportunity to react to Miller’s testimony, he “would have made a very, very strong case that this is not the kind of information that the jury should be allowed to take away.” Tr. Day 3, 277. Lentini, too, opined that Miller simply did not have enough evidence to support his opinion. Tr. Day 1, 58. 38. Had the Spinales sought a voir dire of Miller’s qualifications and the opinions to which he would testify, his testimony would have been limited or excluded. He would not have been permitted to testify to the effect of heat on metal without demonstrated expertise in that area. In light of the poor quality and quantity of the documentation in this case, he would not 21 have been permitted to testify that he saw “no evidence” the fire started in the basement. 39. Titus was subjected to extensive and withering cross-examination on his qualifications. Tr. Tr. VI, 65, et seq. The cross-examination centered on the fact that Titus, unlike the government witnesses, is not primarily an arson investigator, but rather a fire protection engineer, as if that made him less, rather than more, qualified than the government’s witnesses. Titus was ineffectively rehabilitated on this point, even though Jay Spinale knew that Titus’s education, licensing, and experience were far more substantial and rigorous than the largely on-the-job training and experience that informed Domingos’s and Miller’s testimony. Tr. Day 3, 255. As Titus demonstrated at the hearing in this proceeding, his formal education in preparation for licensing as an engineer included numerous graduate-level courses in chemistry, physics, and fire dynamics, as well as mathematics and other engineering-related science courses. Tr. Day 3, 247-49; Ex. 11. In contrast, none of the government’s experts had an engineering license and none had pursued a course of studies in the physical sciences. In every case, the government witnesses who gave expert testimony had criminal justice training and/or a liberal arts education, and their experience was solely in the area of law enforcement (David Domingos, Wayne Miller) or firefighting (Todd Myers). Domingos’s undergraduate degree was in psychology; he also held a master’s degree in criminal justice. Tr. Day 3, 407. He has never studied chemistry or physics at even the undergraduate level. Id. While the government touted their certification as fire investigators by the International Association of Arson Investigators, that certification establishes expertise merely at the level of technician or technologist and is primarily experience-oriented. Tr. Day 3, 252, 253. Certification requires only a certain level of experience and passing a multiple choice examination involving 100-120 questions. Id. 22 Certification does not require the demonstration of an ability to correctly determine the cause and origin of a fire. Id. There is no peer review structure in the CFI certification, there is no control over the quality of the qualifying experience, and to a large extent there is no control over the educational background of a certified applicant. This is in contrast to the requirements for a professional engineering license, for which a peer-reviewed curriculum is established and successful completion of that curriculum is followed by a rigorous, multi-day examination which is reviewed and graded by professional engineers. Id. at 254. In short, there is no comparison between the expertise of a licensed, professional fire protection engineer and a certified arson investigator, who may have little or no education in fire science. Tr. Day 3, 255. Defense counsel should have, but did not, clarify the distinction between a fire protection engineer and a certified fire investigator on redirect examination.12 40. Defense counsel also failed to rehabilitate Titus with respect to the impression left on cross-examination that Titus conceded that the dog “was 100% accurate” and that the literature established the accuracy of canine alerts. See Tr. Tr. VI, 98. Because Titus had made Jay Spinale aware before trial that the technical literature questioned the reliability of canine alerts, Spinale should have known that Titus intended to say, before the prosecutor cut him off, that the literature is widespread in raising the problem of inaccurate alerts, both positive and negative. Tr. Day 3, 255, 258-66. Competent, alert counsel would have corrected the misimpression left by Titus’s incomplete testimony, and there was no conceivable reason to leave 12 NFPA 921, with which Jay Spinale claimed verbatim knowledge at the time of trial, makes clear that a fire protection engineer’s training exceeds that of a certified fire investigator, and may be needed to provide technical assistance to the fire investigator. See Ex. 2, ¶¶ 14.5.1, 14.5.6.1. 23 the jury with the notion that Titus endorsed Billy’s accuracy “100%.” G. 41. Trial Counsel’s Closing Effectively Undercut Hebshie’s Principal Defense: that the Government’s Cause-and-Origin Evidence Did Not Bear Scrutiny In closing argument, defense counsel effectively conceded Domingos’s qualifications to render an opinion on cause and origin: “Now, you heard David Domingos testify, and David Domingos testified that he’s a fire investigator, and, incidentally, I think the guy is a very qualified person, a very straight fellow that told you what he really believes. I don’t quarrel with any of his testimony.” Tr. Tr. VI, 188. Defense counsel similarly endorsed the impressiveness of Billy and handler Doug Lynch. Id. at 189. Ultimately, defense counsel invited the jury to ignore differences between the defense and government theories of the cause and origin of the fire and to rely on their collective common sense. “When it comes to the experts that testified, I suggest to you there’s little difference between what these people say. . . . We don’t need an expert to tell us the fire burns up, heat burns up.” Id. at 198. In effect, after praising the qualifications, reliability, and bravery of the government’s witnesses, defense counsel invited the jury to ignore his own expert’s testimony about cause and origin, thereby seriously undermining Hebshie’s defense. 42. It was unreasonable and incompetent for counsel to abandon its principal defense in closing. No minimally competent lawyer would have done so. There was no possible strategic reason to endorse Domingos and Lynch, nor was there a reason to invite the jury to disregard the substantial differences in the testimony of the defense and government witnesses with the suggestion, essentially, that it was all a wash. H. Hebshie Was Prejudiced by Counsel’s Ineffective Performance 43. The second prong of the Strickland test requires in inquiry into whether there is a 24 reasonable probability that defects in counsel’s performance affected the outcome of the case. In this case, counsel’s ineffective performance unquestionably affected the outcome. Exclusion of any substantial portion of Lynch’s testimony would very likely have affected the jury’s assessment of the government’s case, given its marginal relevance, its incompetence, and its inflammatory emotionalism. Given that the government was never able to identify which among a generic class of thousands of LPD’s was found in the sample, there was a clear failure of proof which the canine evidence obscured. Evidence that the “accelerant-detecting” dog gave a “strong alert” – an entirely subjective assessment with no demonstrated meaning – effectively substituted for reliable, scientifically valid evidence that what the dog alerted to had any tendency to prove that the fire was incendiary and originated at the location of the alert. The canine evidence added nothing to the evidence on either point: it did not establish (1) what, specifically, the chemical was; (2) that the chemical was not incidental to the environment; (3) how the chemical came to be at the spot of the alert; (4) that the chemical was used to start the fire; or (5) that the fire started in the location where the sample was taken. Yet, the government’s extensive focus on the canine evidence invited the jury to infer all of those impermissible deductions. The erroneous admission of this evidence, which was the subject of the longest testimony of the trial, stretching over two days, and the subject of extensive and improper argument by the prosecution in its closing, in all likelihood affected the outcome of this trial. The jury could hardly ignore it, particularly given that defense counsel himself endorsed Lynch and Billy in its closing argument. 44. Exclusion of the laboratory results would have left the government without any evidence, except financial motive, to support an inference that the fire was set. The improper inclusion of the laboratory results in all likelihood affected the outcome. 25 45. The failure to challenge the basis of the overall cause-and-origin opinions, and the qualifications of the government’s opinion witnesses, was unquestionably prejudicial. Had the defense effectively shown that principles of fire dynamics were inconsistent with the government’s analysis of the fire dynamics in this case, a reasonable jury would likely have concluded that the case was not proven. Similarly, if the defense had effectively prevented Todd Myers from testifying incorrectly to the import of the thermal imaging results, the jury would have had little choice but to conclude that the fire was already in the walls when it erupted into open space within the newsstand. 46. As noted in Hebshie’s memoranda filed in support of this petition, this prejudice must be weighed in light of the other deficiencies and errors not addressed herein: the completely irrelevant, improper, and inflammatory evidence that a firefighter suffered a heart attack as a result of the defendant’s conduct and the failure to mount any challenge at all to the mail fraud charge which carried a ten-year on-and-after minimum sentence. 47. The Strickland standard of prejudice is not high. It does not require Hebshie to show even that it is more likely than not that, but for counsel’s errors, he would have been acquitted. But, in this case, the evidence of prejudice is overwhelming. The second prong of Strickland is amply met in this case. Without the cause-and-origin evidence, there would have been insufficient evidence to support the arson charge, and by extension, the mail fraud charge. Had counsel performed effectively, the Daubert challenge might well have resulted in dismissal of all charges. IV. The Government’s Investigation Was So Incomplete and So Poorly Documented that the Indictment Should Be Dismissed Little of the government’s evidence at the hearing was material to the question of 26 ineffective assistance of counsel, but rather was an exercise in face-saving, a justification of an investigation that Lentini described as “miserable.” Tr. Day 1, 29-30. Domingos’s and Marquardt’s defense of the investigation in this case demonstrates how little they really understand the scientific method. While the government witnesses paid lip service to NFPA 921, they followed neither its letter nor its spirit. Domingos appears to have understood the scientific method to mean little more than that an investigator starts from least to greatest areas of damage. He did no vector analysis, collected no data on depth of char, did no arc mapping, event sequencing, analysis of fire dynamics, etc. Tr. Day 1, 79, 89; Tr. Tr. III, 107. Cf. NFPA 921, Chapter 17; Figure 17-2;13 Tr. Day 1, 76. The Phoenix report illustrates how the investigation in this case was marred by tunnel vision and expectation bias, that is, jumping to a conclusion before all evidence was gathered and then limiting the collection of evidence to that which supports the expectation. Tr. Day 1, 56; Ex. 3, ¶ 4.3.8. As early as Monday morning, April 23, at the same time that he expressed a need to see the basement and long before the laboratory sample was analyzed, Domingos told Cronin that the fire started in the convenience store. Ex. 8. Domingos’s conduct exhibits not only tunnel vision on his own part, but also an insensitivity to the problem of contextual bias created when one investigator is informed of the results of another investigator. See generally National Research Council, Strengthening Forensic Science in the United States: A Path Forward 122-23 (2009). Contrary to Domingos’s testimony (Tr. Day 3, 381), NFPA 921 does not require documentation of only the burned areas, but of the “investigation” (Ex. 2, ¶ 15.1.2), including 13 Chapter 17 was rewritten in the 2008 edition. Tr. Day 1, 80, Ex. 2. 27 “evidence of reasonable alternative hypotheses that were considered and ruled out.” Id., ¶ 11.3.5.2. This requirement he utterly failed to meet. In fact, it is clear that Domingos never formulated alternative hypotheses about the origin of the fire; his flippant testimony about alternative causes, including nuclear explosions and lightning, was dismissive of the notion that alternatives must be thoroughly examined. Tr. Day 3, 403. Domingos’s testimony also demonstrated a poor understanding that the purpose of documentation is to support and verify the investigator’s opinions and conclusions. Tr. Day 3, 398; cf. NFPA 921, ¶ 15.1.2. He testified that it was not important to provide sufficient documentation so that a supervisor could check to see whether alternative hypotheses had been properly rejected, because, with three troopers and two local fire department investigators on site, “in a sense we had peer review at the scene.” Tr. Day 3, 400. But, investigators working together to develop group consensus reflected in a single report of investigation are not performing the function of independent review. That function cannot take place without adequate documentation of findings. One effect of Domingos’s poor documentation is that, five and nine years later, testifying under oath, he simply filled in the gaps in the documentation, with demonstrably poor accuracy but little doubt. For example, after Lentini criticized the failure to remove debris from the floor of the newsstand as depicted in the photographs, Domingos claimed that he systematically removed debris from the site so that 90-95% of the area of the newsstand was walkable. Tr. Day 3, 387, 392. That statement is belied not only by the complete lack of documentation of any process of sifting through and removing layers of debris, but by the photographs taken at Cronin’s behest the next morning, which show that the site was still buried in debris. Ex. 9. 28 Another telling example is Domingos’s testimony about the location of the stairwell in the newsstand. Tr. Ex. 7 the diagram of the newsstand floorplan, does not show the stairwell into the cellar. Nor is it mentioned in Domingos’s report. Asked where the stairwell was located, Domingos incorrectly testified that the stairwell was located in front of the partition, and across from the purported point of origin. See Tr. Day 3, 408, 411. In fact, as the government stipulated, the stairwell was located on the right-hand wall, in the back office area, along the wall separating the newsstand from Ro-Ann Jeweler. Tr. Day 4, 522. Another effect of the poor documentation is that other government witnesses who offered opinions either on the cause and origin of the fire or the adequacy of the investigation relied on what Domingos testified to or told them without independent verification. A comparable situation would be if one fingerprint examiner testified that in her opinion the fingerprints were a match, not based on actually reviewing the fingerprints, but based on the testimony of another fingerprint examiner who claimed he did. There is simply no value in such testimony. Domingos attempted to justify the poor documentation of his investigation on the grounds of limited resources. But while he testified that he was hampered by having only limited film at his disposal (Tr. Day 3, 395), the photographs that were taken under his direction (see Ex. 19) include a posed portrait of the troopers with Billy, and several other useless and/or duplicative photographs – a poor use of what Domingos claimed was an extremely limited resource. Domingos’s statement that he did the best he could “on that particular day with the resources that we have available to us” (Tr. Day 3, 385), translates, roughly, to “good enough for government work.” It cannot justify rushing to judgment about how and where a fire started, and, on that basis, bringing charges entailing a 15-year minimum mandatory sentence. 29 In sum, the evidence in this case amounted to no more than Domingos’s ipse dixit, unsound pseudo-science (Myers’s interpretation of the thermal images, the lab report on the sample, and the canine evidence), and photographs which are inconsistent with Domingos’s conclusions. None of the government’s witnesses could explain how a fire starting within the newsstand could possibly, consistent with the laws of physics, have resulted in the fire damage to the grate or the header or lintel of the stairwell to the cellar. None of them explained how a fire starting within the area of the shop managed to go sideways to enter the pipe chase, while leaving most of the plywood intact. None of them could satisfactorily explain how a fire starting in the newsstand could possibly have resulted in white spots on all four walls after the fire was extinguished at the suspected place of origin. None of them credibly explained how the damage to the B wall and the areas not damaged along that wall are consistent with the fire having started within the room. See Exs. 4, 5. As Lentini explained, had the fire started where the government witnesses posited, one would expect to see more damage at that location than is evident in the photographs. The fire would have spread out horizontally at the area of the shelf, and the thin plywood siding on the walls would have be much more heavily burned. Tr. Day 1, 51. If anyone ever went into the cellar, and it appears unlikely that anyone did, then the absence of fire observed there means that the fire shot up through the pipe chases and vacant spaces in the walls and burned at higher levels. It was imperative for the investigators to carefully examine the cellar ceiling below the suspected point of origin along the B wall. The “V” pattern was two feet wide at the floor, indicating that the point of the V was below the newsstand floor level. Tr. Day 1, 37. All of the phenomena observed in Exs. 4 and 5 are consistent with the fire having come through the pipe 30 chase and are not consistent with the fire having started in the shop. Tr. Day 1, 38-39. While the government more or less conceded that Domingos’s investigation did not follow the letter of NFPA 921, the defense of the investigation seems to be: nowhere does NFPA 921 say how bad an investigation has to be in order to render it unreliable. This defense holds no water. No manual setting forth a consensus of good practice would or could prescribe the point at which an investigation is so compromised that the results are invalid. What is clear is that, as much as the prosecution would like to cabin the deficiencies in the investigation as narrowly as possible, they were fundamental and pervasive and they compromised the entire cause-and-origin determination. What the failure to document the site amounts to in this case is spoilation of evidence. The state fire investigators had control of the scene and released it without preserving sufficient evidence. Given that it is now impossible to reconstruct the fire scene, to perform arc mapping, to examine the floor for burn patterns, to examine the underside of the floor on the B side of the newsstand, to see the condition of electrical components in the basement under the jewelry store and under the newsstand, or to collect additional samples for laboratory analysis, Hebshie is deprived of the ability to conduct an investigation necessary for his defense. Moreover, it is plain that the government cannot withstand a Daubert challenge to the entire cause-and-origin determination. Accordingly, the indictment in this case should be dismissed and Mr. Hebshie released from custody forthwith. 31 CONCLUSION For the foregoing reasons and for those raised in Hebshie’s 2255 petition, and the memoranda and affidavits submitted in support of the petition, he respectfully submits that the petition be granted, the indictments dismissed, and that he be released from custody. Respectfully submitted, JAMES G. HEBSHIE By his attorney, /s Jeanne M. Kempthorne Jeanne M. Kempthorne BBO No. 267410 7 Bott’s Court Salem, MA 01970-3101 (978) 745-4673 Date: August 20, 2010 CERTIFICATE OF SERVICE I hereby certify that I have this day caused the foregoing Proposed Findings and Rulings to be served on counsel of record by means of the CM/ECF system. /s/ Jeanne M. Kempthorne Date: August 20, 2010 32