UNITED STATES of America, Plaintiff-Appellant, v. Boyce Mark GARRETT; Larry Don Keith; Kenneth Vernon Rydeen; Jimmy Dale Sullivan; Michael Dean Woodard; Steven Craig Finklea; Austeen Harris Keith; Dale Zane Keith; Glenn Dale Wilcox, Defendants-Appellees, 238 F.3d 293

DeMOSS, Circuit Judge: I. INTRODUCTION In this interlocutory appeal, the government seeks review of an order of the district court, which was entered the day of trial and which excluded 25 of its witnesses in a complex, multi-defendant conspiracy case involving the alleged adulteration of milk. The district court itself recognized that exclusion of these witnesses was tantamount to a dismissal of some of the charges against the moving defendants.

This appeal boils down to one relatively uncomplicated issue; that is, whether the district court abused its discretion by imposing the rather draconian sanction of excluding the government’s witnesses from trial for discovery violations which the court itself found not to have been made in bad faith. For the reasons discussed below, we find that the district court abused its discretion. We, therefore, VACATE the district court’s sanctions order excluding the government’s witnesses and REMAND this matter for further proceedings. II.

BACKGROUND The underlying cause of action giving rise to the criminal indictments in this case involves a complex and massive, long-term conspiracy in which more than 30 dairy farmers and milk transporters allegedly added water to milk shipments along various and overlapping dairy routes in order to increase both the weight and value of milk shipments.

The indictment described how Associated Milk Producers, Inc. (“AMPI”), which is a marketing corporation for dairy farmers, operates a milk processing plant in Sulphur Springs, Texas, to which drivers it hired would bring milk from individual dairy farms along each driver’s specified route. The route drivers had the responsibility of measuring the quantity of milk received from each dairy farm and had to take samples of each shipment received from each farm before commingling the milk in the tanker truck.

Once delivered to the processing plant, the milk was to be tested under standards put forth by the Milk Marketing Administration (“MMA”), which is a subdivision of the U.S. Department of Agriculture. The government’s theory was that various individuals, including route drivers, added water to the milk trucks along delivery routes. To prove its case, the government intended to rely on both scientific data, that is, sample test composition reports for the milk tanker trucks, and testimony from cooperating witnesses.

These witnesses were drivers and other AMPI staff members who knew of or were aware of the defendants’ schemes to water down the milk. Most of the testifying witnesses whose testimony was ultimately excluded by the district court were defendants who pleaded guilty to one count of the indictment in exchange for dismissal of the remaining counts against them and for consideration of a downward sentencing departure in light of their willingness to cooperate and testify truthfully in the government’s case against the remaining defendants.

Along with substantive violations of specific milk adulteration statutes, specifically 21 U.S.C. §§ 331(a), 331(a)(2), the indictments charged substantive mail fraud violations and various conspiracies to: 1) defraud the government by obstructing milk regulation; 2) violate the milk adulteration statutes; and 3) commit mail fraud. The initial indictment, entered on July 15,1998, charged 29 defendants with these various milk adulteration-related offenses.

The case was initially set for trial on September 28, 1998, but was rescheduled for November 30, 1998, then for January 11, 1999, and then again for March 15, 1999 (with a February 12th deadline for discovery). On January 13, 1999, a superseding indictment was returned naming four additional defendants, and on March 3, 1999, a second superseding indictment was returned. Four days before the March 15th trial date, the district court held a hearing on various motions, and the following day, March 12th, it entered an order continuing the trial once again to April 5th.

The matters addressed in the March 11th hearing dealt with allegations that the prosecutor declined to produce letters written to people who were not expected to testify at trial (“target letters”), urging them to admit their involvement to get the benefit of cooperation at sentencing, and that the prosecutor also declined to provide investigators’ notes of interviews with or questionnaires as to approximately 125 people, in some of which various defendants denied any involvement, a position which was inconsistent with their pleas and which, therefore, constituted impeachment material that should have been disclosed.

In its March 12th order, the district court directed the government to produce within five days of that order, copies of “any [target] letters from government counsel or its agents attempting to secure testimony from or against any person who will be testifying in the government’s case-in-chief....” The district court also continued the trial until April 5th to permit production of these materials. On March 16th and 17th, the government produced some materials in response to the March 12th order.

On March 31st, and in response to a defense motion that all correspondence with any individuals, not just correspondence related to testifying witnesses, be produced, the district court removed the “case-in-chief’ limitation of its March 12th order and ordered that all such target letters be produced by April 2nd. On March 23rd, all of the defendants-appellees had also moved jointly for relief under Brady claiming that the government had withheld numerous categories of exculpatory materials, including the letters referenced in the March 12th order.

On April 1st, the district court denied the defendants’ motion, noting that none of the materials cited by the defendants constituted Brady material for which the government had an affirmative duty to disclose. The district court did note, however, that the investigators’ notes and screening questionnaires were to be produced as “Brady impeachment material” because they revealed potentially inconsistent statements made by the various defendants. Yet, no defense counsel ever denied having been told by the prosecutor about these items and their content.

Late on April 2nd, the government produced documents to the defense as directed by the district court’s March 31st modification of its March 12th order. The government supplemented this production on the morning of April 5th (the trial date) with a stack of documents (8 inches thick), a good portion of which was dupli-cative of previously provided materials. Defense counsel complained to the district court about the tardy production, and while some conceded that a brief continuance would suffice to take the sting out of the delayed production, others requested a dismissal based on discovery violations.

On April 5th and 6th, the district court held additional hearings on the defendants’ various motions to dismiss or for sanctions regarding the government’s alleged failure to timely produce discovery materials, and it ultimately struck 25 of the government’s witnesses on the afternoon of the 5th. The district court ordered excluded from trial any witness as to whom a target letter was required to be produced by the court’s March 12th order (requiring production by March 17th), but as to which such letter was tardily produced.

The court noted that after the March 11th hearing, it had determined, and the government should have been aware that, such target letters were Brady materials whieh it had an obligation to produce. The district court did not sanction the government’s failure to timely disclose the materials ordered produced after April 2nd under the March 31st modification of the March 12th order (i.e., the addition of any target letters to non-testifying witnesses and the addition of the investigators’ notes).

A written order to the effect of the sanctions orally ordered was entered on April 9th, and in it, the district court stated “[t]he court does not question the government’s good faith.” The district court went on to state: even assuming the government’s untimely production ultimately would not have adversely impacted the defense’s trial strategy, it nonetheless unquestionably adversely impacted the organized and efficient preparation for trial by defense counsel and, for that matter, the Court.

Neither the defense nor the Court should be forced to continue to suffer the government’s last-minute production of documents and the resulting motions and hearings the untimely productions have caused, particularly when those untimely productions have come on days immediately preceding, and even the day of, trial.

The government has timely filed this interlocutory appeal of the district court’s sanctions order, arguing that the district court abused its discretion by imposing a sanction more severe than was necessary to effect compliance with discovery orders and by failing to weigh all of the factors required by this Circuit’s precedent. III. STANDARD OF REVIEW We review a district court’s imposition of sanctions for discovery violations for an abuse of the district court’s discretion. See United States v. Katz, 178 F.3d 368, 372 (5th Cir.1999).

The government concedes that the district court’s discretion is “admittedly broad.” However, notwithstanding this broad discretion, we have consistently held that a district court, when considering the imposition of sanctions for discovery violations, must carefully weigh several factors, and if it decides such a sanction is in order, it “should impose the least severe sanction that will accomplish the desired result — prompt and full compliance with the court’s discovery orders.” United States v. Sarcinelli, 667 F.2d 5, 7 (5th Cir. Unit B 1982); see also Katz, 178 F.3d at 372. IV.

DISCUSSION As we noted in Sarcinelli and Katz, a district court exercising its discretion and considering the imposition of sanctions for discovery violations should consider the following factors: 1) the reasons why disclosure was not made; 2) the amount of prejudice to the opposing party; 3) the feasibility of curing such prejudice with a continuance of the trial; and 4) any other relevant circumstances. See Katz, 178 F.3d at 371 (citing United States v. Bentley, 875 F.2d 1114, 1118 (5th Cir.1989)).

And as noted above, in fashioning any such sanction, the district court should impose only that sanction which is the least severe way to effect compliance with the court’s discovery orders. Id. The government relies heavily on a decision from the Eleventh Circuit extolling the principles of Sarcinelli, which, though not controlling, is instructive. In United States v.

Euceda-Hernandez, 768 F.2d 1307 (11th Cir.1985), the court noted that by suppressing governmental evidence in lieu of granting a continuance or recess, “a trial judge may achieve a speedier resolution ... and reduce his docket, but he does so at the expense of sacrificing the fair administration of justice and the accurate determination of guilt and innocence.” Id. at 1312.

In the government’s view, the district court’s striking of 25 of its witnesses, which the district court itself recognized could have the effect of eviscerating the criminal indictment, was “tantamount to a dismissal of charges ... [and constituted] an undeserved windfall to parties who were duly indicted based on probable cause to believe they had committed federal crimes.” The government urges that if it violated any discovery order at all, it acted in good faith, and that there was no measurable prejudice to the defendants which could not have been cured by a short continuance.

Finally, the government urges that less severe sanctions, such as personal sanctions against the prosecutor, could have achieved the goal of compliance with discovery orders. The defendants counter that the district court did not even go far enough, as some sought dismissal of the indictment as a sanction. And likewise, the defendants do not feel that any less severe sanction would suffice to ensure that the government would comply with the district court’s discovery orders.

The defendants contend that the record shows a pattern of misconduct by the government that makes the sanctions chosen by the district court mild, and they further contend that the sanctions were a valid exercise of the district court’s power to sanction discovery violations “as a prophylactic and punitive measure.” To determine whether the district court abused its discretion, we must evaluate the exercise of discretion in light of our precedent requiring that the district court fully and thoughtfully addressed each of the Sarcinelli factors noted. A.

Reasons for non-disclosure We first consider the government’s reason for not timely producing target letters for the 25 excluded witnesses. The government explained to the district court that its failure to more timely provide the target letters resulted from the fact that the letters were in a separate binder that had been overlooked during initial disclosures. As noted above, the district court explicitly noted that the government did not violate its discovery orders in bad faith and that its late production was the result of an unintentional mistake.

Indeed, no improper motive was attributed to the government’s tardy production. The government relies on a decision from the D.C. Circuit in which the court held that such a severe sanction as suppression of evidence would rarely be appropriate when the trial court finds the violation not to have been made in bad faith and where a less dramatic remedy, such as a continuance, would mitigate any prejudice. See United States v. Marshall, 132 F.3d 63, 70 (D.C.Cir.1998).

We note also that in our own decision in Sarcinelli, we found the prosecutor’s complete failure to provide discovery at all to be contumacious, but nevertheless, not deserving of the harsh sanction of exclusion tantamount to a dismissal where a less severe sanction such as jailing the prosecutor or granting a continuance was available.

The district court’s own finding that the government’s tardy disclosure was not in bad faith militates against the imposition of a sanction so severe as to effectuate a dismissal of the charges against certain defendants, especially where as discussed below, other, less severe sanctions were available to mitigate against the minimal prejudice suffered by the defendants in this case.

The defendants rely on what they characterize as a pattern of disclosure abuses, which indicates that the government’s untimely disclosure was an intentional move designed to overwhelm the defendants at the last minute so as to prevent them from being able to utilize the disclosed target letters.

Despite the defendants’ characterization of their 21 separate requests for discovery throughout the many continuances of this case, the defendants overlook the abundance of materials that were in fact timely produced by the government and the fact that, as the district court noted, the failure to timely provide the target letters as to the 25 witnesses was the result of an unintentional mistake. We conclude that the district court’s own findings are dispositive of the good faith issue and that the reason for non-disclosure was a mistake made in good faith. B.

Prejudice to the defendants The second Sarcinelli factor we must consider is whether the defendants were unduly prejudiced by the tardy disclosure. The district court assumed that the late production would not prejudice the defendants’ trial strategy, but it found that tardy disclosure so close to the commencement of the trial “adversely impacted the organized and efficient preparation for trial by defense counsel and ... the Court,” because it required the filing of motions and the scheduling of hearings.

We note that even though trial was set to commence on April 5th, the district court had already scheduled, at one defense attorney’s request, a recess from Tuesday the 6th until Thursday the 9th, to accommodate the attorney’s scheduled appearance for oral argument before this Circuit. The government properly notes that the prejudice referred to in Sarcinelli is prejudice to the defendants’ substantial rights, that is, injury to their right to a fair trial, and that prejudice does not encompass putting trial preparation into minor disarray. See United States v.

Webster, 162 F.3d 308, 336 (5th Cir.1998), cert. denied, 528 U.S. 829, 120 S.Ct. 83, 145 L.Ed.2d 70 (1999); see also United States v. Neal, 27 F.3d 1035, 1050 (5th Cir.1994). As we noted in our decision in United States v. Martinez-Perez, 941 F.2d 295, 302 (5th Cir.1991), the question of prejudice is whether the defendant had time to put the information to use, not whether some extra effort was required by defense counsel.

Additionally, even if the district court determines that the information was disclosed too late to be put to effective use, the court must also determine that the lack of information created a reasonable probability that the result would have been different. See Kyles v. Whitley, 514 U.S. 419, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995).

In order to determine in this case whether the result would have been different, the district court should have considered all of the materials that were in fact produced to evaluate whether the target letters would have made a difference; however, when the government attempted to make a record of the cumulative nature of the target letters to show that the prejudice, if any, was minimal, the district refused to grant it an opportunity to do so because of the government’s “admission” of a discovery violation and the district court’s assumption that damage to the defendants’ case was not necessary to support the sanction.

We have repeatedly held that no prejudice exists when suppressed or newly discovered evidence is cumulative. See, e.g., United States v. Lowder, 148 F.3d 548, 551 (5th Cir.1998); Allridge v. Scott, 41 F.3d 213, 217-18 (5th Cir.1994). The precise materials that were deemed to be a discovery violation in this case were 23 simple target letters to 23 witnesses, a draft immunity agreement with a 24th witness, and a target letter with a proposed offer agreement for a 25th witness.

Among the information already in the hands of the defendants at the time these materials were tardily produced were the following: as to 7 of the 25, the defense had other target letters with the same message; as to those 7 plus 12 more, the defense had threatening and coercive correspondence following up on target letters; and as to 21 of the 25, the defense knew that they had pleaded guilty and struck deals with the government and these 21 defendants’ plea agreements, including the dismissal of charges and the stated possibility of a § 5K1.1 motion, were known to the defense.

Quite simply, it should have come as no surprise to the defense that the pleading defendants had previously received target letters encouraging them to plead guilty. As we noted in Webster, where a defendant is impeached with his plea agreement and his memorialized hope for a reduced sentence, additional information regarding anticipated favors from the government in exchange for cooperation is cumulative impeachment material that is not prejudicial if untimely or undisclosed. See Webster, 162 F.3d at 337-38.

We find it highly unlikely that the failure to have the undisclosed materials would have hindered the defense’s ability to impeach the 25 excluded witnesses regarding their prior deals with the government, which deals might call in to question their motivation for testifying against the remaining defendants. In our view, the district court completely overlooked the additional evidence that renders the target letters cumulative and, thus, a minimizing factor of the prejudice suffered by the defendants. C.

Curing prejudice imth a continuance The district court itself acknowledged in its April 1st order denying the exclusion of various witnesses on Brady grounds that continued violations of its discovery deadlines and scheduling orders, in addition to “wreaking havoc” on the defense’s and the Court’s ability to efficiently prepare for trial, “might require additional continuances of the trial date.” In doing so, the district court implicitly recognized that a continuance was a viable and likely consequence of tardy disclosure.

Additionally, most of the defense attorneys conceded to the district court that if there was going to be a continuance, it would only need to be for two or three days. Furthermore, as the sanctioned materials were scheduled to be produced by 5:00 p.m. on Friday,- April 2nd, only two days prior to the actual disclosure on Monday the 5th (the scheduled day of trial), and as the district court had already planned to recess the trial from Tuesday the 6th through Thursday the 8th, we conclude that a brief continuance of several days would not have impacted either the district court’s

schedule or the defendants’ ability to efficiently prepare for trial. In light of the absence of bad faith on the part of the government, the minimal amount of substantive prejudice because of the cumulative nature of the tardily disclosed materials, and the availability of a much less severe sanction than striking witnesses with the effect of eviscerating the government’s case, we find that the district court could most certainly have eliminated the minor prejudice with either a brief delay or a less severe sanction. D.

Other relevant factors Sarcinelli lastly requires that the district court also consider those additional matters which are relevant to a determination of whether sanctions are appropriate. The government presents one main, but persuasive point regarding this factor, which is that allowing such a harsh sanction to stand in these circumstances essentially obliterates its case against individuals who were duly indicted based upon probable cause to believe they committed crimes against the government.

As we have stated, a district court “exceeds the proper bounds of its power to order dismissal of an indictment ... when it fails to consider whether less extreme sanctions might maintain the integrity of the court without punishing the United States for a prosecutor’s misconduct.” United States v. Welborn, 849 F.2d 980, 985 (5th Cir.1988)(citing Sarcinelli, 667 F.2d at 6-7).

Here, we conclude that the exclusion of these 25 witnesses, with the effect of eliminating or substantially diminishing the government’s case against the defendants-appellees, was an excessive sanction and an abuse of the district court’s discretion, especially where a brief continuance would have cured any prejudice and other sanctions were available to ensure that the prosecutor would comply with the district court’s discovery orders. V.

CONCLUSION Based upon our full consideration of the Sarcinelli factors, which should have guided the district court’s decision on sanctions, we conclude that the district court abused its discretion in excluding 25 of the government’s witnesses. The government acted not in bad faith, the prejudice to the defendants was minimal in light of the cumulative nature of the untimely disclosures, and any prejudice could have been cured with a brief continuance. For these reasons, we VACATE the district court’s order of sanctions and REMAND this matter for such further proceedings as are appropriate.

VACATED and REMANDED. .By November 1998, as a result of numerous guilty pleas, only five defendants remained in the case (Appellees Garrett, Larry Don Keith, Rydeen, Sullivan, and Woodard). The original indictment was relumed in July 1998. In August 1998, fourteen defendants pleaded guilty; in September, four more pleaded guilty; in October, two more pleaded guilty; in November, three more pleaded guilty; and one more defendant was dismissed from the indictment by virtue of his wife having entered a guilty plea.

Each of these defendants agreed to cooperate with the government, and all bargained to have their sentencing hearings postponed until after dispositions had been reached with respect to all defendants charged in the indictments. Each was presumably hoping for a 5K1.1 downward departure based on substantial assistance. . These additional four were Finklea, Austeen and Dale Keith, and Wilcox. Combined with the five remaining defendants as of November 1998, those being Garrett, Larry Don Keith, Rydeen, Sullivan, and Woodard, see supra note 1, we have our nine named appellees. . Brady v.

Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963). . We note that this is inconsistent with the fact that, in its April 1st order, the district court explicitly held that the target letters, as one of the categories of materials for which the defendants’ decried a failure to disclose, were not Brady materials. Thus, the government cannot be said to have been on notice that such letter were even Brady materials until the district court entered its order on March 12th requiring disclosure within five days thereafter. .

This assumption contravenes the second Sareinelli factor requiring a full consideration of actual prejudice.