UNITED STATES DISTRICT COURT DISTRICT OF CONNECTICUT
VERNON HORN,
Plaintiff,
V. No. 3:18-cv-1502(RNC)
CITY OF NEW HAVEN, ET AL.,
Defendants. _______________________________ ______________
MARQUIS JACKSON,
V. No. 3:19-cv-388(RNC)
Defendants.
RULING AND ORDER Plaintiffs Vernon Horn and Marquis Jackson bring these consolidated actions under 42 U.S.C. § 1983 and state law against the City of New Haven, former New Haven Police Department Detectives Leroy Dease, Petisia Adger and Daryle Breland, and State of Connecticut firearms examiner James Stephenson. The actions arise from a murder that occurred in 1999 during an early
morning robbery of a 24-hour convenience store in New Haven. The plaintiffs were convicted of the murder and other offenses after a jury trial based on the testimony of Steven Brown, who pleaded guilty and
testified that the plaintiffs were primarily responsible for the robbery-murder. Brown testified that Horn was armed with a pistol that looked like a Beretta and began firing it immediately upon entering
the store. Stephenson testified that based on his analysis of the ballistics evidence, the murder weapon could have been a Beretta. The plaintiffs’ convictions
were vacated in 2018 in light of newly discovered evidence. The charges were then dismissed. The plaintiffs have always claimed that they never
met or spoke with Brown prior to their arrests for the robbery-murder, and they continue to maintain that they are actually innocent. They claim that their wrongful convictions and lengthy imprisonment were caused by
the Detectives, the City, and Stephenson - the Detectives because they engaged in investigative
misconduct; the City because it was deliberately indifferent to the Detectives’ violations of the plaintiffs’ rights as suspects in the investigation; and Stephenson because he failed to disclose
exculpatory information to the trial prosecutor in violation of Brady v. Maryland, 373 U.S. 83 (1963). The Amended Complaint advances thirteen claims in
all. Discovery having been completed, all the defendants have moved for summary judgment on all the claims. The motions filed by the Detectives have been
granted in part and denied in part. See Horn v. City of New Haven, No. 19-cv-388, 2024 WL 1261421 (Mar. 19, 2024)(ruling on Detectives’ motion for summary judgment on federal claims); Horn v. City of New Haven, No. 19-
CV-388, 2024 WL 1342762 (Mar. 29, 2024)(ruling on Detectives’ motion for summary judgment on state law claims). And the City’s motion has also been granted in part and denied in part. See ECF 330, 334. This
memorandum addresses Stephenson’s motion. For reasons that follow, the motion is granted.
I. On January 24, 1999, at about 3:25 a.m., three masked gunmen burst into the Dixwell Deli in New Haven.
The first to enter sprayed five or six bullets from a 9 mm pistol in the general direction of the cash register, killing Caprice Hardy, a customer, and wounding Yousif Abbey, an employee.
Four days later, the New Haven Police Department sent cartridge casings, bullets and bullet fragments
recovered from the scene to the State Police Forensic Science Laboratory and requested that they be analyzed. Stephenson, an experienced firearms examiner at
the Lab, received the request submitted by the NHPD. His duty was to determine what kind of gun or guns might have fired the bullets recovered from the crime
scene. For this purpose, firearms examiners use the General Rifling Characteristics Database (GRC Database) maintained by the F.B.I. and distributed to firearms examiners throughout the country. The goal of querying the GRC Database is to
identify firearms that could have fired bullets recovered from a crime scene. A query includes the caliber of a bullet, the number of lands and grooves on the bullet (impressions made by the firearm’s barrel on
the bullet), and measurements of the land widths and groove widths on the bullet, in one thousandths of an inch.
The examiner also includes a margin of error (or “tolerance range”) for either side of measurements: so if the land width is 50, and the margin of error is +/-
2 one thousandths of an inch, the database will search for results with a land width of 48 to 52. The Database then provides a list of firearms manufacturers and models consistent with the information submitted by
the examiner. In selecting a margin of error, an examiner makes a judgment call based on his training and experience
and his measurements of the physical evidence. The objective is to set the margin of error at what might be called the Goldilocks point – narrow enough to avoid
getting too many firearms in the response but not so narrow as to produce an unhelpful result. The margin of error that makes sense under the circumstances depends on the measurements of the physical evidence.
With several projectiles that all have the same measurements, a small margin of error may produce a usable result. With different measurements, a larger margin of error may be necessary.
Many firearms examiners use +/-5 one thousandths of an inch as their normal default range. This is in
accordance with tolerance ranges approved by the F.B.I. GRC Manual. It is also consistent with accepted standards of practice in the field.
Stephenson measured the land and groove impressions on the bullets and bullet fragments and queried the GRC Database to determine what kind of gun could have fired the bullets. In formulating his
query, he chose a margin of error (or “tolerance range”) of +/-2 one thousandths of an inch based on his training, experience, judgment and measurements of the
physical evidence. He “thought that [+/-2] was the right margin of error to use under the circumstances.” It “made sense . . . because [he was] pretty confident based on the physical measurements that the actual land
and groove width of the firearm that fired the[] bullets would be close to 69 and 105.” In response to the query, the computer system
produced a printout – called a GRC Report - containing a list of manufacturers and models of firearms that could have been used to fire the projectiles recovered
from the scene. Beretta was not among them. On February 4, 1999, Stephenson prepared a standard report in the format used by the Firearms
Section of the State Police Forensic Science Lab. The report described the evidence submitted and the results of the firearms examination. The report was signed by Stephenson and another firearms examiner in the
Forensic Science Lab. Stephenson sent the report to the NHPD. The report stated: The bullet and bullet fragments are consistent with being 9mm caliber. They may have been fired from but not limited to a self loading pistol manufactured by Calico, FEG, Browning, Heckler & Koch, Hungarian, Kassnar, Norinco or Walther.
During Stephenson’s time working as a firearms examiner in the State Forensic Lab, it was normal, standard practice to transmit a signed report to the requesting agency setting forth the examiner’s conclusions without also transmitting work materials such as GRC documents, worksheets, or other notes or research material. An examiner would provide these materials to the prosecutor in advance of a criminal
trial if a specific request was made by the prosecutor or defense attorney. In other jurisdictions, it was standard practice to affirmatively provide exculpatory information to the prosecutor without a request. But
that was not the practice in Connecticut in 1999. After Stephenson’s report was sent to the NHPD, the Dixwell Deli investigation led to the arrest of
Steven Brown. In a post-arrest interview, Brown admitted that he participated in the Deli robbery and
claimed he was coerced into doing so by the plaintiffs. He said that Horn had a 9 mm Beretta that he began firing as soon as walked in the door.
Approximately one year later, Assistant State’s Attorney Gary Nicholson was preparing to try the robbery-murder case against the plaintiffs. On reviewing Stephenson’s report, he noticed that Beretta
was not listed as a possible match with the ballistics evidence. In view of Brown’s statement that Horn was armed with a Beretta, Nicholson decided to follow up
with Stephenson to find out if it was possible a Beretta could be the murder weapon. Nicholson called Stephenson and directed him to
conduct a further examination of the ballistics evidence to see if a Beretta would fit within the parameters of the evidence. Nicholson had no idea what Stephenson would need to do to answer this question.
For example, he was not “familiar with the concept of margin of error in general rifling characteristics.” He simply wanted Stephenson to check whether the
ballistics evidence that he had already examined was consistent with potentially coming from a pistol manufactured by Beretta. How Stephenson checked was up to him.
In response to Nicholson’s question, Stephenson queried the GRC Database again. He did not re-examine the evidence before doing so.
Stephenson submitted two queries with new inputs: for the first query, he used the same margin of error
of +/-2 he used before but added “9MM LUGER” and “PISTOL”; for the second, he increased the margin of error to +/-4. In response to both queries, Berettas appeared as possible matches. The second query also
identified four other potential manufacturers that did not appear in response to his query in 1999.
Based on the results of these new queries, Stephenson called Nicholson and told him the murder weapon could be a Beretta. He did not tell Nicholson how he arrived at that finding, and Nicholson did not ask. Ordinarily, Stephenson’s reports to Nicholson,
whether written or oral, did not go into how he arrived at his findings. Stephenson’s expert witness has testified that in
like circumstances, he would have informed the prosecutor of the steps he had taken in response to the prosecutor’s question. In other words, he would have told the prosecutor, in substance, “If I query the
database with the tolerance range I think is right – the one I used before - I don’t get a Beretta. It’s only when I change it that I get a Beretta.”
Had Nicholson been aware that Stephenson “widened the margin of error which added a whole bunch of new manufacturers to match the bullet, not just the
Beretta, he would have turned that information over to defense counsel.” As it was, he simply informed them that he had been in touch with Stephenson to find out if the murder weapon could be a Beretta, and Stephenson
had said yes. Nicholson called Stephenson as a witness to
testify at the criminal trial. On the day he was scheduled to testify, Stephenson brought his entire work file to the courtroom, where it would be available to the prosecutor, defense counsel and the judge. It
was standard practice for Stephenson and other firearms examiners at the State Lab to bring their entire work file (including GRC documents, worksheets or other notes or research material) to court proceedings and to
make it available to the prosecutor, defense attorney, or judge, in case anyone wanted to see the file or any of its contents, and to be cooperative in answering
questions concerning the file materials. Horn’s defense counsel, Leo Ahern, was an experienced criminal defense lawyer at the time of the
Dixwell Deli trial, having previously tried approximately 10 to 12 homicide cases, and a few more than that involving the discharge of firearms. He had a number of other cases with Stephenson before the
Dixwell Deli case. In Ahern’s experience, it was Stephenson’s practice to always bring his work file to
court. Ahern has no recollection of whether he did or did not ask Stephenson to see any materials in his file in the Dixwell Deli case. Anytime he did ask Stephenson, he was allowed to see the materials. On no
occasion did he ask to see Stephenson’s research materials and been told no. He saw Stephenson’s “ballistics report” before the trial. Typically, he would get the firearms examiner’s ballistics report –
the examiner’s signed formal report – in discovery prior to trial. He believes he may have had some familiarity with GRC Reports in 1999. But he cannot
remember whether he ever saw that type of document in any case prior to 1999. Nor can he recall one way or the other whether he saw any GRC Reports in advance of Stephenson’s testimony in the Dixwell Deli case. He
hardly remembers Stephenson’s testimony and does not recall questioning him on the stand. Jackson’s counsel, Michael Moscowitz, was also a
highly experienced criminal defense lawyer at the time of the Dixwell Deli trial. He recalls speaking with
Stephenson about the case, either by phone prior to the trial or in the courthouse before Stephenson took the stand. They spoke about Stephenson’s anticipated testimony. Moscowitz does not recall whether or not he
looked at Stephenson’s work file. But he is sure Stephenson brought the file to court because he always did. Prior to the Dixwell Deli case, Moscowitz had many cases in which Stephenson was a witness for the
State. Moscowitz never found Stephenson to act in anything other than the most upright, responsible, cooperative manner. In no case did Stephenson not
allow Moscowitz to look at the material in his work file. In Moscowitz’s experience, Stephenson was “extremely cooperative” and “highly professional.” He would show Moscowitz documents and discuss them,
answering any questions Moscowitz might have. Moscowitz does not remember one way or the other if he looked at any GRC Reports in the Dixwell Deli case. At the criminal trial, Stephenson testified on
direct examination by Nicholson that when he first analyzed the ballistics evidence to determine which firearms could have been used in the robbery-murder, he came up with a list of numerous manufacturers. The
examination continued: Q Now, a couple of weeks ago, did you have – did I have occasion to have a phone call with you? A Yes, you did. Q Did I make a request for you to check some additional information? A Yes, you did. Q And based upon that request, did you come up with any other manufacturers that also fit within the parameters of this ballistics evidence? A Yes I did. Q Can you tell us what that was? A The other firearms manufacturer was Beretta.
Stephenson’s expert has testified that this is “not what [he] would have testified to” because Stephenson only identified the one manufacturer the prosecutor was looking for, which was “not full disclosure.” Stephenson has testified: “I have no
reason to understand why I didn’t say anything further. I don’t know why I didn’t say anything further.” On cross examination by Mr. Horn’s counsel,
Attorney Ahern, the following exchange occurred: Q Mr. Stephenson, to reiterate briefly, your original report is dated February 4, 1999. Is that correct? A Yes, sir, it is. Q And then there were inquiries made of you, am I right, from Attorney Nicholson representing the State. Is that correct? A That’s correct. Q Subsequent to the report? A Yes. Q When were those inquiries made? A I believe it was February 15th of this year. Q February 15 of the year 2000? A That’s correct. Q More than a year later than the original report. Is that correct? A That’s correct. Q Now, let’s see, what exactly was he asking you to do? A He asked if a Beretta semiautomatic pistol possibly would fall within the ranges of the rifling characteristics. Q And did you generate a report of any kind as a result of the inquiries? A Just notes that I kept for myself, but there was no report generated through the department. Attorney Ahern does not recall whether or not at the next recess he asked Stephenson to look at the notes he referred to in his answer. His view is that the GRC Reports would constitute “notes” of the type Stephenson referred to during his testimony on cross- examination. So, if he did ask to see Stephenson’s
notes, the GRC Report would be the type of document he would expect Stephenson to show him.
Plaintiffs contend that Stephenson’s response to Horn’s counsel on cross-examination that he did not generate a report of any kind in 2000 was “a lie” because in truth he did generate a GRC Report showing
his use of a margin of error of +/-4. According to the plaintiffs, “Stephenson falsely denied that he had created any reports in 2000 because he did not want the defense to know he changed the margin of error.” The plaintiffs’ assertion that Stephenson lied on
the stand is premised on their position that the term “report” means the result of a GRC search, which Stephenson refers to as a “printout.” Stephenson says the term “report” refers to the examiner’s signed
report identifying the evidence examined and setting forth the examiner’s conclusions. What matters here is not how the term “report” is
used and understood today, or even how it was generally used and understood in 2000. What matters is how the participants in the Dixwell Deli case used and
understood the term. I think a reasonable juror would have to find that Attorney Ahern, in his line of questioning, used the
term “report” to refer to the “ballistics report” he was accustomed to receiving in every case, not a GRC Report, which he has no memory of seeing in any case prior to 1999.
As recounted above, Ahern’s line of questioning begins with a question about Stephenson’s “original report dated February 4, 1999,” that is, his signed
report. The questioning then moves on to the inquiries that were made to Stephenson “[s]ubsequent to the report,” “more than a year later than the original report.” Attorney Ahern then asks, did you generate a
report of any kind as a result of these inquiries?” to which Stephenson replies “there was no report generated through the department.”
The colloquy plainly shows that both Ahern and Stephenson were using the term “report” to refer to a report that Stephenson would prepare and sign – the
type of report that would be generated through the department, not a GRC report generated by a computer. Any uncertainty about this is dispelled by Attorney Ahern’s deposition testimony. At his deposition he was
asked a series of questions about the documents he was used to seeing in cases involving firearms examinations. He testified that he distinguished between “ballistics reports,” on the one hand, and GRC
reports on the other: the former are the firearms examiner’s “formal report” or “report”; the latter are
part of the examiner’s “notes.” II.
Following their release from prison, the plaintiffs brought these actions seeking compensation for their wrongful convictions and lengthy
incarceration. With regard to Stephenson, they alleged that (1) he took it upon himself to alter his 1999 query of the GRC Database in order to be able to give Nicholson the desired opinion that the murder weapon
could be a Beretta; (2) did not disclose to Nicholson the changes to the 1999 query that enabled him to provide that opinion; (3) testified at the criminal trial that the murder weapon could have been a Beretta;
and (4) again failed to disclose how he come to that conclusion.
In response to the complaint, Stephenson filed a motion to dismiss on the grounds of qualified and absolute immunity. Judge Meyer denied the motion and Stephenson took an interlocutory appeal. On appeal, he argued that (1) qualified immunity protected him
against the plaintiffs’ claim because it was not clearly established in 1999 that firearms examiners had an obligation under Brady to turn over exculpatory evidence to the prosecutor; and (2) he was entitled to
absolute immunity with regard to his work in 2000 because he did it at the prosecutor’s direction. The Court of Appeals disagreed. It concluded that based on the facts alleged in the complaint, Stephenson was not
entitled to either qualified or absolute immunity. III.
The parties’ briefs present the following arguments:
Prosecutorial immunity Stephenson renews his argument that he is entitled
to prosecutorial immunity because the work he did in 2000 was at Nicholson’s direction. The plaintiffs respond that the Second Circuit “resoundingly rejected” Stephenson’s reliance on prosecutorial immunity “on the same facts” and its ruling is binding here. The
plaintiffs castigate Stephenson for ignoring the Second Circuit’s ruling: “The remarkable dishonesty of Stephenson’s argument reveals its utter frivolousness. Stephenson’s motion for summary judgment on the basis
of absolute prosecutorial immunity lacks any merit, is sanctionable and must be denied.” Witness immunity
Stephenson contends that the plaintiffs cannot use his criminal trial testimony to prove their Brady claim
and that, without it, the claim cannot succeed. In response, the plaintiffs state that Stephenson’s “failure to disclose the favorable information embodied in the [second] 2000 [GRC] Report existed before his
trial testimony and is independently actionable under § 1983.” The idea that his false trial testimony immunizes him from liability is, they submit, ”absurd.”
Sufficiency of the Evidence Stephenson contends that the plaintiffs cannot
prevail on their Brady claim because they cannot establish that he (1) withheld information from Nicholson or (2) that he did so with the culpable mental state necessary to support a verdict in their
favor. In response, the plaintiffs emphasize that Stephenson does not challenge their ability to prove the other elements of their Brady claim - “favorability to the defense” or “materiality.” As to the elements
he does put in issue - “suppression” and “culpable mental state” – they say a jury could find that he deliberately concealed information from Nicholson and
defense counsel because, although he made his entire file available to them in the courtroom prior to taking the stand, he failed to proactively tell them how he came to his finding that the murder weapon could be a
Beretta. In addition, the plaintiffs say, a jury could reasonably find that Stephenson committed perjury to prevent Nicholson and the defense counsel from finding out that he changed the query of the GRC Database he used in preparing his original report, which he
considered “right” at the time, in order to get a result that included a Beretta. IV.
Before turning to the issues raised by the parties’ arguments, it is necessary to clearly describe
the Brady material at issue. In their opposition brief, the plaintiffs state that Stephenson faces liability under Brady because he failed to disclose that the murder weapon could not have been a Beretta,
and that proper analysis of the ballistics evidence would exclude a Beretta.1 But the plaintiffs do not have a firearms examiner of their own who says either
1 See Pls’ Joint Mem. of Law in Opp. to Def. Stephenson’s Mot. for Summ. J., ECF 284, at 24 (“That the murder weapon could not have been a Beretta, when cooperator Steven Brown claimed Vernon Horn killed the victim with a Beretta, plainly qualifies [as material under Brady]”; id. at 23 (“Stephenson offers no argument against either the exculpatory value or the materiality of the fact that the murder weapon could not have been a Beretta if the analysis was done ‘right’”); id. at 20 (“Plaintiffs seek to hold Stephenson liable for failing to disclose to the prosecutor that the murder weapon could not be a Beretta when analyzed the ‘right’ way, and could only be a Beretta when reanalyzed the wrong way, as reflected in the February 1999 and February 2000 GRC Reports.”). of these things. Nor do they have any other evidence
permitting a reasonable finding that the murder weapon could not have been a Beretta or that proper analysis would exclude a Beretta. From this conspicuous and unexplained lack of evidence, I infer that the murder
weapon could have been a Beretta, as Stephenson testified at the criminal trial. In their joint brief, the plaintiffs refer to
“exculpatory information contained in the 2000 GRC Report.” In this and other parts of the brief, they say that Stephenson should have disclosed to Nicholson
that he generated a report listing Beretta as a possible manufacturer and model after increasing the margin of error from +/-2 to +/-4. See ECF 284 at 29. They state that it was “Stephenson’s obligation to
disclose this information to the prosecutor so the prosecutor could disclose it to the defense.” Id. at 25. In light of this, I understand the plaintiffs to
be claiming that what Stephenson was required to say to Nicholson is, in substance, the following: You have asked me to check whether a Beretta could be the murder weapon, pointing out that the nonexclusive list in my previous report does not include a Beretta. In order to respond to your question, I have conducted two queries of the GRC Database that differed from the one I used last year: first, I used the same margin of error of +/-2 but added the words Luger and pistol.2 The result was a list that included Beretta. I then did a second query using a margin of error of +/-4, which is larger than the +/-2 margin of error that I used before but less than the default margin of error of +/-5 approved by the F.B.I.3 Again, the result included Beretta. Based on these results, my answer to your question is yes, it is possible the murder weapon could be a Beretta. As I understand the plaintiffs’ theory, Stephenson’s failure to tell Nicholson how he arrived at his answer that the murder weapon could have been a Beretta violated Brady because the information was (1) favorable to the defense (2) Stephenson intentionally
2 The plaintiffs do not reckon with the first query Stephenson did in 2000 using a margin of error of +/-2, but it is undisputed that it was done. 3 The plaintiffs do not dispute that the F.B.I. uses a default margin of error of +/-5. withheld the information to prevent its use at trial,
and (3) had the information been disclosed there is a reasonable probability the outcome of the criminal trial would have been different.
A. Returning to the parties’ arguments, the first
issue is whether prosecutorial immunity protects Stephenson against liability. Prosecutorial immunity extends to individuals who
act under a prosecutor’s direction in performing functions closely tied to the judicial process. Horn v. Stephenson, 11 F.4th 163, 173 (2d Cir. 2021). On the interlocutory appeal, the Court of Appeals said:
We need not reach the question of whether Stephenson generated the 2000 GRC Report in furtherance of the prosecutor’s advocacy function because there is no allegation that Nicholson requested a new report. Horn pleaded simply that the “Assistant State’s Attorney Nicholson . . . called . . . [and] asked Stephenson whether it was possible the murder weapon could have been a Beretta.” Then, “[o]n February 15, 2000, Stephenson generated a new General Rifling Characteristics Report, this time manipulating the report to increase the margin of error to +/-4.” Even if we conclude that adjusting the margin of error constituted prosecutorial advocacy, the complaint nowhere alleges that Nicholson asked, much less instructed, Stephenson to create a new GRC Report using a larger margin of error. Moreover, Horn affirmatively alleged that Nicholson never saw the 2000 GRC Report prior to trial, which further suggests that it was not created at his request. 11 F.4th at 173-74. The plaintiffs contend that, given the language in the Court of Appeals’ opinion, for Stephenson to win on prosecutorial immunity, Nicholson had to specifically direct him “to create a new GRC Report using a larger margin of error.” 11 F.4th at 173. Since Nicholson did not do so, Stephenson is not immune. The Second Circuit did not have the benefit of Nicholson’s testimony when it ruled on the interlocutory appeal. The question for me is whether the Court would still reject prosecutorial immunity for
Stephenson based on the record as it presently exists. I cannot agree with the plaintiffs that the answer is resoundingly clear. Nicholson has testified that he directed
Stephenson to check if a Beretta would fit within the parameters of the ballistics evidence. That was the direction he gave. He had no idea what Stephenson would need to do to find the answer, but he directed
him to check and get back to him. Nicholson has further testified that it was Stephenson’s standard practice to provide his findings without going into what he did to arrive at them. Viewed in this context,
Nicholson was directing Stephenson to get back to him with an answer, yes or no, with no expectation that Stephenson would explain the methodology he used to get
the answer. Suppose Nicholson told Stephenson, “I direct you to check to see whether the Beretta is a possible fit
with the ballistics evidence by using whatever legitimate techniques firearms examiners use for this purpose and get back to me as soon as possible with a simple yes or no.” In complying with such an order,
would Stephenson be acting under the direction of Nicholson in performing a function closely related to
trial preparation? As I read Nicholson’s deposition testimony, that is a fair summary of what actually occurred. Nicholson
needed Stephenson’s help to check whether a Beretta could be the murder weapon because he lacked the necessary knowledge and experience himself. Trial was just a week or so away. He told Stephenson he needed
an answer and directed him to check. He did not specifically instruct him to use the same margin of error he used before, or any other margin of error. He
could not give such a specific instruction because he “was not familiar with the concept of margin of error.” In responding to Nicholson’s request, Stephenson initially used a query with margin of error of +/-2,
then a query with a margin of error of +/-4. As far as the record shows, these queries were legitimate techniques that a competent firearms examiner acting in good faith would use in the circumstances to determine
whether in fact the murder weapon could be a Beretta. If a prosecutor directs an expert to help him
prepare for trial on an issue with respect to which the prosecutor needs the benefit of the expert’s knowledge and experience, is it the law that the expert cannot share the prosecutor’s immunity unless the prosecutor
tells the expert exactly what to do? If so, the immunity ostensibly available to experts working under the direction of a prosecutor to help prepare for trial often will be no immunity at all.
Even so, the issue of prosecutorial immunity cannot be decided favorably to Stephenson as a matter
of law. At a minimum, the Second Circuit’s ruling requires Stephenson to demonstrate that the analysis he performed in response to Nicholson’s direction was legitimately called for by the direction he received
and within the scope of his role assisting Nicholson prepare for trial. Whether that is the case cannot be determined without factfinding by a jury. If a jury were to find that Nicholson directed Stephenson to use
accepted techniques of analysis and Stephenson did so, prosecutorial immunity could be available to him
consistent with the Court of Appeals’ ruling on the interlocutory appeal. If a jury were to find that the techniques he used were illegitimate or not within the scope of Nicholson’s direction, then immunity would not
apply. B.
The next issue is whether Stephenson is entitled to summary judgment based on absolute witness immunity. Trial witnesses have absolute immunity from any § 1983
claim based on the witness’s testimony. Briscoe v. Lahue, 460 U.S. 325, 332-33 (1983)(police officer witnesses have absolute immunity like any other witness). See Rehburg v. Paulk, 566 U.S. 356, 369
(2012)(grand jury witness enjoys same immunity as witness at trial).
The Second Circuit has instructed that when a defendant in a § 1983 case claims absolute immunity for his testimony, it is necessary to determine whether the plaintiff can prove the elements of the claim without resorting to the testimony. If the claim cannot be
proven without the testimony, the defendant enjoys absolute immunity. See Coggins v. Buonora, 776 F.3d 108, 113 (2d Cir. 2015)(dealing with grand jury testimony).
The plaintiffs contend that Stephenson’s reliance on absolute witness immunity has two flaws: his false testimony at the criminal trial does not immunize him
against the Brady claim; and his criminal trial testimony is admissible on the elements of suppression and materiality under Brady.
I agree with the plaintiffs’ argument that the Brady claim is not precluded by witness immunity because Stephenson’s “failure to disclose the favorable
information embodied in the February 2000 GRC Report ‘existed before’ his trial testimony.”
However, that Stephenson’s criminal trial testimony may be used to prove the Brady claim without depriving him of the protection of witness immunity is incorrect. Coggins directs me to consider whether the plaintiffs can prove the claim without resorting to
Stephenson’s trial testimony. If the claim cannot be proven without his testimony, he is absolutely immune. Whether the plaintiffs can prove the elements of the Brady claim at issue here – suppression and culpable
mental state - without his criminal trial testimony is the subject of the next section. C.
Stephenson argues that a jury could not reasonably find that he suppressed anything in his work file,
including the GRC Reports, because he brought the entire file to court and made it available to Nicholson and defense counsel, just as he and his fellow examiners always did to satisfy their Brady
obligations. The plaintiffs do not deny that it was standard practice in Connecticut for a firearms examiner from the State Forensic Lab to bring his work file to court and make it available to the other trial
participants. Nor do they allege that he deviated from this practice in this instance. Instead, they argue that the standard practice and Stephenson’s adherence
to it are legally irrelevant. In their view, Stephenson is liable because he did not affirmatively disclose to Nicholson the methodology he used in responding to Nicholson’s directive, specifically, that
his second query used a larger margin of error than the one he used the year before. In other words, Stephenson had a duty under Brady to proactively raise this matter with Nicholson, and this duty could not be
satisfied merely by making the work file available to him to peruse if he wished.
The plaintiffs are correct that in the Second Circuit, an officer has an obligation to disclose Brady material to a prosecutor without waiting for a request, and it does not matter whether defense counsel could
have obtained the material with reasonable effort. See Horn v. City of New Haven, 2024 WL 1261421, at *10. Accordingly, this element of the plaintiff’s claim can be proven without resorting to Stephenson’s trial
testimony. Stephenson contends that a jury could not
reasonably find that he failed to disclose his use of the increased margin of error with the culpable state of mind required for liability.
In response, the plaintiffs point to Stephenson’s trial testimony. They state that a jury could find that he committed perjury to prevent defense counsel from learning about his use of the increased margin of
error. The “lie” they point to is his denial that he prepared a new report of any kind when in fact he generated a GRC Report using an increased margin of
error. Plaintiffs submit that “[a] rational jury could conclude that this blatant lie under oath shows Stephenson’s intent to hide from the defense what he did in February 2000.”
In addition, the plaintiffs point to Stephenson’s “cover up” while on the stand of the other manufacturers whose names came up when he went back and
checked the ballistics evidence at Nicholson’s request to see if a Beretta was a possibility. Instead of mentioning the names of all the manufacturers that came
up, he mentioned only Beretta, the one the prosecutor was looking for. The plaintiffs contend that “[a] jury could reasonably infer from Stephenson’s perjury and cover-up his bad faith and intentional violation of
Brady.” Indeed, in the plaintiffs’ view, Stephenson is invoking witness immunity to avoid losing on this very
element of the Brady claim. As they put it, “of course, the real reason why [he] wants his testimony to be excluded: His lies and dissembling on the stand are
devastating evidence of his malign intent.” Based on the plaintiffs’ submissions, it therefore appears that to prove Stephenson acted with the
culpable mental state required for liability, they intend to rely primarily, if not solely, on his allegedly perjurious trial testimony. They do not point to anything else. To withstand Stephenson’s
motion on this point, though, they need to demonstrate that they can prove this element of their claim without resorting to Stephenson’s trial testimony. That is
what Coggins says in so many words. Because they have not done so, the motion will be granted based on absolute witness immunity.
D. The next issue is whether Stephenson is protected
by qualified immunity. “The qualified immunity doctrine protects government officials from suits seeking to impose
personal liability for money damages based on unsettled rights or on conduct that was not objectively unreasonable.” Tenenbaum v. Williams, 193 F.3d 581, 595-96 (2d Cir. 1999). “[W]hether an official
protected by qualified immunity may be held personally liable for an allegedly unlawful action generally turns on the ‘objective legal reasonableness’ of the action,
assessed in light of the legal rules that were ‘clearly established at the time the action was taken.’” Messerschmidt v. Millender, 565 U.S. 535, 546 (2012) (quoting Anderson v. Creighton, 483 U.S. 635, 639 (1987)). “The objective reasonableness test is met –
and the defendant is entitled to immunity – if ‘officers of reasonable competence could disagree’ on the legality of the defendant’s actions.” Lennon v. Miller, 66 F.3d 416, 420 (2d Cir. 1995)(quoting Malley
v. Briggs, 475 U.S. 335, 341 (1986). Under this test, qualified immunity will be available unless no officer of reasonable competence could have made the same choice in similar circumstances. “In short, qualified
immunity “protects all but the plainly incompetent or those who knowingly violate the law.” Messerschmidt, 565 U.S. at 546 (internal quotations omitted).
Stephenson contends that he is entitled to qualified immunity because a reasonable official in his position in 2000 could think he had no obligation to
proactively disclose to Nicholson the GRC Reports of the queries he conducted after receiving Nicholson’s directive. He submits that a reasonable firearms examiner in his position could think that the GRC
documents did not constitute Brady material, and that he would fulfill the duties of his office by following
the standard practice of making his work file available to the prosecutor in court. I agree. The duty to disclose Brady material applies to
evidence favorable to the accused that “rises to a material level of importance.” Kyles v. Whitley, 514 U.S. 419, 438 (1995). It would not have been obvious to a firearms examiner in Stephenson’s position that
the GRC Reports generated by his queries of the GRC Database in 2000, viewed together, fit this description. The plaintiffs argue that the second of
the two GRC Reports was exculpatory because it showed Stephenson’s use of a margin of error of +/-4, which was higher than the margin of error of +/-2 he used the year before. In making this argument, the plaintiffs
ignore the first of the two GRC Reports, which shows Stephenson’s use of a margin of error of +/-2. Given that the first GRC Report, derived from a query using a margin of error of +/-2, yielded Beretta as a
possibility, the exculpatory value that the plaintiffs attach to the second one is dubious. In the
circumstances, a reasonable firearms examiner could think that, in view of the first GRC Report, the second GRC Report did not rise to a material level of importance.
In addition, even if a reasonable officer in Stephenson’s position could think the second GRC Report was somehow exculpatory, he could be confident that his
duty to obey the law required no more than that he bring his work file to court and make it available to the prosecutor in accordance with standard practice.
The record establishes that it was the standard practice for firearms examiners at the State Lab to send signed reports to the requesting agency, like the NHPD, and to forward any additional information –
exculpatory or otherwise - only in response to a specific request. In the absence of such a request, it was standard practice for the examiner to bring his work file to court and make it available to the
prosecutor at that time. Under this standard practice, a firearms examiner had no obligation to review his
work file before going to court and proactively provide Brady material to the prosecutor. The record establishes that this standard practice
was a well-established feature of the criminal justice process, well-known to prosecutors and defense counsel alike, including Nicholson and the plaintiffs’ defense counsel. All the witnesses who are competent to
testify uniformly agree that this is the way things worked in Connecticut. No witness has testified that the legality of the practice was ever questioned.
There is no evidence that Stephenson and his colleagues were ever advised that the practice might violate their disclosure obligations under Brady.
As shown by the parties’ Local Rule 56 Statements, it is undisputed that Stephenson subjectively believed he was following standard operating procedures and that, in doing so, he subjectively believed he was
acting properly. Nonetheless, the plaintiffs argue that he is not entitled to summary judgment based on qualified immunity. They argue that viewing the
evidence most favorably to them, the standard practice among firearms examiners in 2000 actually required Stephenson to proactively disclose to Nicholson any exculpatory information in his possession, including
raw notes or other research materials. Since he failed to do so, he is not entitled to qualified immunity. The plaintiffs’ argument is based on practice in
other jurisdictions, not Connecticut. That the standard practice in Connecticut was as described above is not genuinely disputed. Moreover, the plaintiffs
have not shown that it would have been obvious to a reasonable firearms examiner in Stephenson’s position that the second GRC Report had exculpatory value rising to a material level of importance.
V.
Accordingly, the motion for summary judgment is hereby granted. So ordered this 1st day of April 2024 _______/s/ RNC_______________ Robert N. Chatigny United States District Judge